CA2301435C - Method for reducing congestion in packet-switched networks - Google Patents
Method for reducing congestion in packet-switched networks Download PDFInfo
- Publication number
- CA2301435C CA2301435C CA002301435A CA2301435A CA2301435C CA 2301435 C CA2301435 C CA 2301435C CA 002301435 A CA002301435 A CA 002301435A CA 2301435 A CA2301435 A CA 2301435A CA 2301435 C CA2301435 C CA 2301435C
- Authority
- CA
- Canada
- Prior art keywords
- network
- congestion
- packet
- endpoint
- operational condition
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Expired - Fee Related
Links
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L12/00—Data switching networks
- H04L12/02—Details
- H04L12/14—Charging, metering or billing arrangements for data wireline or wireless communications
- H04L12/1485—Tariff-related aspects
- H04L12/1496—Tariff-related aspects involving discounts
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L12/00—Data switching networks
- H04L12/02—Details
- H04L12/14—Charging, metering or billing arrangements for data wireline or wireless communications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L12/00—Data switching networks
- H04L12/02—Details
- H04L12/14—Charging, metering or billing arrangements for data wireline or wireless communications
- H04L12/1485—Tariff-related aspects
- H04L12/1489—Tariff-related aspects dependent on congestion
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L47/00—Traffic control in data switching networks
- H04L47/10—Flow control; Congestion control
- H04L47/11—Identifying congestion
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L47/00—Traffic control in data switching networks
- H04L47/10—Flow control; Congestion control
- H04L47/28—Flow control; Congestion control in relation to timing considerations
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L65/00—Network arrangements, protocols or services for supporting real-time applications in data packet communication
- H04L65/80—Responding to QoS
Abstract
The present invention permits a network service provider to detect an operational condition - such as congestion - in a packet-switched network and to allevia te such congestion by providing customer incentives to avoid use of the network. The detection mechanism triggers an incentive such as the modification of the user's acces s charges and the customer can be immediately notified of either the occurrence of the congestion or of information regarding the incentive. Usage of the network during congested periods can be deterred by imposing additional access charges during such periods - similarly, customers can be given a discount to encourage usage during periods of low congestion. An incentive schedule can be tailored to dynamically change the usage patter ns of the customers of the network to accommodate the operational conditions in the network. The present invention has application in the Internet, where a detection/notification mechanism, for example, can be implemented in a network node such as a muter or in a network endpoint such as a client machine or a proxy or mail server.
Description
METHOD FOR REDUCING CONGESTION Ill PACKET-SWITCHED
NETWORKS
Field of the Invention The present invention relates to communication networks and, more particularly, to methods and systems for reducing congestion in a packet-switched or hybrid network.
background of the Invention With the explosive growth of today's information superhighway have come the inevitable traffic jams. Congestion is a serious problem today on the Internet, a worldwide system of computer networks using packet-switching technology to transfer 1o messages between computers. Packet-switching protocols such as the Transmission Control Protocol/Internet Protocol (TCP/IP) divide messavges into packets which travel along a path in the network that can be varied as conditions in the network change.
Specifically, TCP/IP, as currently implemented in the Internet, routes packets independently of each other, utilizing its best efforts without any specific concept of a 15 "connection". Accordingly, in the Internet, there is little :notion of "quality of service", no notion of guaranteed throughput, and no notion of bounded transmission delay.
Current implementations of TCP/IP rely on packet loss as an indicator of congestion in the network. As the network experiences congestion, data flowing through a network router becomes bottlenecked in a queue until tl~e queue overflows and packets 2o are lost. Load reduction is accomplished by utilizing a well-known "congestion avoidance" algorithm first described by Van Jacobson in 1988. See "Congestion Avoidance and Control", V. Jacobson, ACM SIGCOMM-88, August 1988, p. 314-29;
"TCP Slow Start, Congestion Avoidance, Fast Retransmit:, and Fast Recovery Algorithms", W. Stevens, RFC 2581 (revision of RFC 2001). In what is coined a "slow 25 start", a TCP source begins inserting packets into the network by starting with a minimal congestion window, allowing at most one unacknowledged packet in the network.
Each time an acknowledgement (ACK) is received, the congestion window is enlarged exponentially until a first threshold is reached or until a packet is dropped.
If the first threshold is reached, the TCP source continues to enlarge the congestion window linearly until either a second threshold is reached - or until a packet is dropped.
Upon the timeout of a retransmit timer (thereby indicating a dropped packet), the TCP source reduces the transmission rate and "backs ofd' to its minimal window, with the goal of allowing the network to reach some form of equilibrium.
As traffic on the Internet increases and more applications are run which are 1o sensitive to the delay caused by dropped packets (e.g. stre;aming audio and video), proposals have emerged to add some form of explicit congestion notification (ECN) to TCP. See "TCP and Explicit Congestion Notification", ACM Computer Communication Review, V. 24 N. 5, ~ctober 1994, p. 10-23; "A Proposal to add Explicit Congestion Notification (ECN) to IP", I~.I~. Ramakrishnan and Sally Floyd, RFC 2481. For example, 15 a network router with a queue nearing an overflow, rather than merely waiting for a packet to drop, can transmit a signal (in the form of a special bit in the packet's header) to indicate the presence of network congestion. The receiver's acknowledgment packet passes the notification on to the sender, which in turn slows down its transmission rate.
These methods of controlling congestion by signaling for a reduction in 2o transmission rate, however, do not address the root problem - namely, insufficient transmission capacity to support the explosive growth in the number of users demanding access to the Internet at the same time. Short of increasing the capacity of the network, these methods of addressing congestion can make performance degradation more gradual, but they cannot prevent it altogether. It would be preferable to establish a system that 25 reduced congestion in the network by affecting the network usage habits of the people accessing the Internet, and thereby directly addressing the problem of overrunning the capacity of the network.
Summary of the Invention The present invention permits a network service provider to detect an operational condition - such as congestion - in a packet-switched network and to alleviate such congestion by providing customer incentives to avoid use of the network. The detection mechanism triggers an incentive such as the modification of the user's access charges and the customer can be immediately notified of either the occurrence of the congestion or of information regarding the incentive. Usage of the network during congested periods can be deterred by imposing additional access charges during such periods -similarly, customers can be given a discount to encourage usage during periods of low congestion.
1o An incentive schedule can be tailored to dynamically change the usage patterns of the customers of the network to accommodate the operational conditions in the network. The present invention has application in the Internet, where a detection/notification mechanism, for example, can be implemented in a network node such as a router or in a network endpoint such as a client machine or a proxy or mail server.
In accordance with one aspect of the present invention there is provided a method, for use by a network access service provider to alleviate an undesired operational condition in the network, said method comprising the steps of: detecting the undesired operational condition in the network; detecting current usage of the network by a customer; and providing a financial incentive to the customer to cease current use of the 2o network and attempt use at another time when the undesired operational condition is not present.
These and other advantages of the invention will be apparent to those of ordinary skill in the art by reference to the following detailed description and the accompanying drawings.
Brief Description of the Drawings Fig. 1 is a block diagram of a packet-switched network illustrating an embodiment of the present invention.
Fig. 2 is a diagram showing the movement of illu:~trative packets in a packet-switched network between a sender and a receiver as a function of time.
Fig. 3 is a block diagram of a packet-switched network and a proxy server illustrating another embodiment of the present invention.
Fig. 4 is a block diagram of a packet-switched network illustrating another embodiment of the present invention.
Fig. 5 is a diagram of a user display with a popup window in the upper right hand corner of the display showing information regarding congestion in a packet-switched network.
Detailed Description The present invention is illustrated with reference to Fig. 1 which shows a packet-switched network 100 having numerous packet-switching nodes 110 to 119 connecting endpoints 101 and 102. Endpoint 101 and nodes 110 to 115 are assumed to be under the control or supervision of a network service provider; the remainder of the network is 1o assumed to be controlled or maintained by other providers or entities. The network service provider provides access for users to the network for an access charge. For example, where the network 100 is the Internet, an Internet Service Provider ("ISP'~ such as AT&T WorldNetT'" provides access to the Internet for its customers. In the case of the Internet, 110 to 115 represent routers and endpoints, 101 and 102 can be client machines, 15 servers, proxy servers, mail servers, news servers, etc. Fig. 1, of course, is a simplification as a typical communication network would encompass other network elements that would be apparent to one of ordinary skill in the art.
Furthermore, although the discussion below focuses on service providers, one of ordinary skill would easily recognize that the present invention applies equally to other network entities such as, for 2o example, corporate networks that utilize charge-back schemes.
In accordance with the present invention, operational conditions such as congestion in the network are detected in a network node 110 to 115 and/or at an endpoint 101 in the network under the control of the network service provider.
Upon the detection of the condition, whether at a router or an endpoint, the present invention 25 generates incentive information, such as billing records reflecting a reduction or an increase in the access charge paid by a particular user. This information can be relayed to a billing server or some other billing apparatus for processing. A
notification mechanism permits the user to receive notice of the incentive, either by notifying the user of the presence of the operational condition (e.g. congestion) or of the incentive information (e.g. the modified access charge).
The specific detection mechanism will depend on the particular operational condition in the network sought to be detected as well as the protocols that can be used in the network to signal the condition. For example, each node 110-115 is customarily equipped in a packet-switched network with a large number of buffers for storing packets in anticipation of routing or awaiting availability of an output link. With regard to packet congestion, such symptoms develop first in the node's buffers or queues, as the buffers become filled and unavailable to store incoming packets. Thus, a router knows that the 1o Internet is getting congested because its buffer queue for some output link is too long or is getting too long. Some routers today utilize a mechanism called Random Early Detection ("RED") which signals the presence of congestion as it develops by dropping packets when the average queue length exceeds some threshold - rather than waiting for the queue to actually overflow. See RFC 2309, which is incorporated by reference 15 herein.
Where endpoint 101 is a sender of packets across the network, see Fig. 2, it will also be aware of the congestion developing in the router's buffers. Endpoint 141, using TCP/IP, expects to receive an ACK 202 after transmitting a packet 201 through the network. Failure to receive an ACK signifies that the packet has been dropped by some 2o router between it and the destination endpoint 102. Where, however, endpoint 101 is a receiver of packets (and endpoint 102 is, accordingly, the sender), the situation is a bit more subtle. The TCP process layer at endpoint 101, as it receives the packets 201, 203, etc. sent by 102, knows the order in which to reassemble the packets based on a sequence number in the received packets' headers. Endpoint 101, thus, expects to receive the 25 packets in a certain order and can infer the dropping of a packet by looking for "holes" in the packets' sequence numbers. An out-of sequence packet, especially if there is a significant delay before the hole is filled, in general indicates that the expected packet has been dropped due to congestion. This method does not guarantee absolute detection of every dropped packet since, for example, packet loss will be invisible to the receiver if the trailing packet/packets in a sequence are dropped. Nevertheless, the method statistically provides good detection of dropped packets, especially for long transmissions (which is the situation a network service provider would be the most concerned about).
Alternatively, the endpoint could use duplicate packets as an indicator, although the method would not be expected to be as good as a method based on detecting a hole and a timeout.
The situation is simplified if the network has ECN capabilities. In that case, where the muter experiences congestion in its buffers (whether by a buffer overflow or by RED), it sets a "Congestion Experienced" ("CE") bit in the packet header of packets 1o from ECN-capable transport protocols. See RFC 2481. The receiver of the packet detects the CE bit and sets a "ECN-Echo" flag in the header of a subsequent ACK packet sent back to the sender. Endpoints 101 and 102 are thus quickly notified of the congestion condition in a router and can react accordingly.
The above detection mechanisms have been described with respect to the Internet 15 protocol suite although, as noted above, the present invention works with applications and protocols other than reliable data transfer over TCP (as well as non-TCP/IP networks such as Ethernet, hybrid networks, etc.). For example, the instant methods of congestion-based incentives work with multicast communication as well as unicast communication.
Consider a multicast audio application that runs over RTP/(TDP (the Real Time Protocol 2o over the User Datagram Protocol). In multicast audio, a sender transmits a stream of packets containing audio samples to multiple receivers. These applications do not require that every audio sample be reliably delivered, but they do require some reliability in order to maintain acceptable audio quality. RTP in particular uses sequence numbers to order packets and detect losses at each receiver. A lost RTP packet signals congestion just as a 25 lost TCP packet does. RTP packets can also arrive at a receiver, in principle, containing an ECN (Explicit Congestion Notification) signal placed there by the network.
Furthermore, RTP receivers send periodic reports back to RTP senders. Senders use these reports to monitor communication quality and possibly adapt their behavior when there are problems. RTP reports include loss information, just as TCP
acknowledgements do. Report packets can also carry ECN-Echo signals back to the sender. Therefore, both implicit and explicit congestion signals can be used in the context of the present invention as akeady described above.
To see how the present invention can be implemented in a real-world setting such as the Internet, it is necessary to understand the typical operating environment.
About 70 to 75% of the traffic in the Internet today utilizes the Hypertext Transfer Protocol (HTTP), i.e., Web page retrieval. Furthermore, most congestion in the case of HTTP will occur on user-bound packets, since that is the direction of most Web traffic.
Some users connect directly to the Web server of interest; others go through what are 1o known as proxy servers. Often operated by network service providers, a proxy server acts as an intenmediary between a user terminal and the Internet to provide caching services. By caching frequently-requested pages, a proxy server can reduce the bandwidth necessary for the provider's own connection to the outside world.
The present invention, then, should handle the case of both direct and proxy connections to the 15 outside world, especially (but not exclusively) for Web traffic.
First, consider the case of a user of a proxy server. With reference to Fig.
3, proxy server 301 receives a request from the user 300 for an Internet service (e.g. a Web page request from server 302) which the proxy server forwards through the Internet 305 using its own IP address - assuming the request passes filtering requirements and cannot 2o be satisfied by the internal cache. When the proxy server receives the requested page, it relates it to the original request and forwards it back to the relevant user.
A proxy server, as a receiver of packets from outside Web servers, can use the present invention to detect congestion in the network and attribute it to the customer who requested the specific Web page. It is also a sender on the provider's own network; this, too, can be noted 25 appropriately. The proxy server is thus perfectly positioned to detect and charge for requests over congested networks. T'he same analysis holds for operator-provided e-mail and, to a certain extent, news servers maintained by the operator. When sending or receiving mail on behalf of a given customer, operational conditions internal and external to the network can be detected, noted, and billed accordingly. (It should be noted, however, that there could be philosophical problems with regard to the application of the present invention to the receipt of unsolicited mail during periods of congestion; see the discussion below on service-level congestion.) Suppose, though, that the user does not utilize a proxy server and is connecting directly to the Internet, whether for Web-browsing or something else. See Fig.
4. In this case, the end systems that detect the congestion most easily are the customer's machine 401 and some endpoint 402 not under control of the network operator. While the former is ideally placed to notify the user, it clearly cannot be trusted to generate charge records for purposes of billing. Under certain circumstances, it may be possible to gather data to reliably, such as when the customer utilizes a tamper-resistant modem supplied by the network operator. Otherwise, the routers 410 to 41 S in Fig. 4 operated by the network service provider should be used to detect congestion and gather the data necessary to generate the appropriate billing records.
The only case not covered, then, is when congestion occurs, but on a router not 15 under the control of the local network operator, i.e. routers 416 to 419.
Arguably, this is not a matter of local concern, since the local network operator is not paying for the bandwidth. There is a situation of interest, however - namely, when the router detecting the congestion is at the other side of a comparatively slow link between the local network operator and an upstream service provider. There are two possible ways to overcome the 2o problem: First, the local operator can provide the routers for both ends of the slow link.
At the upstream end, a fast link can connect to the local provider's router;
this, then, reduces to the previous case. Alternatively, by contractual arrangement the upstream provider can detect and record congestion on behalf of the local operator.
This situation can be generalized. Network service providers can detect congestion attributable to other 25 customers of other network operators, and notify and bill them appropriately. A special-purpose "congestion indication protocol" can be utilized to pass the information between network operators. Accordingly, all users of an upstream provider can be charged for the congestion that they cause, rather than trying to attribute the problem to individual users.
Clearly, notifying the users of the incentive is important: if users do not realize there is a pmblem in the network (and a surcharge/reward), they will not modify their behavior at appropriate times. Direct notification from the network provider is straightforward where the present invention is implemented as a user process running on a direct endpoint capable of detecting congestion. With other configurations, other mechanisms for notifying the customer can be utilized. For example, when the user is in contact with only some local server, service-specific mechanisms can be utilized. When the customer is accessing Web pages through a proxy, a Java or Javascript applet can be sent to the user in the first Web page retrieved which, in turn, displays to the user the relevant notifications regarding congestion and incentive information such as the effect on access charges. Similarly, the mail retrieval protocol could be modified to send appropriate information to the customer when accessing a mail server to check the user's e-mail (or create a new e-mail).
The notification can be in the form of a window or screen "popup" on the user's display. A small window could be displayed on the user terminal indicating the level of congestion in the network in some visually intuitive and appealing form such as a speedometer. See Fig. 5. Moreover, other forms of notification can come from pre-emptive detection mechanisms. In commonly assigned U.S. Patent No. 5,870,557, "Method for Determining and Reporting a Level of Network Activity on a 2o Communications Network Using a Routing Analyzer", the disclosure of which is incorporated by reference herein, a method is disclosed for periodically analyzing the congestion along routes from an access provider's entry point to the Internet to a set of desired Web sites. Round trip times and packet loss information collected from use of the "traceroute" command are utilized to identify congested links. The transit characteristics are compiled and analyzed to provide a user with useful information about congestion along routes to the Web sites of interest and in order to warn users to avoid traffic to some Web sites at congested periods of time. Such preemptive mechanisms can be utilized with the present invention to provide a customer with prior information on the operational condition of the network before incurring any charges for usage of a congested link. Accordingly, a customer can be prompted before deciding to access the network in a manner that will generate any incentive information such as a modification to a customer billing record.
Moreover, it should be apparent from the above example that the notification mechanism need not be coupled to the network's detection and incentive mechanisms.
Self notification mechanisms can be utilized by the user. As long as users have received some notice of the general contours of the incentive, perhaps at registration, they can utilize whatever local detection/notification application scheme on their own computer terminal they choose to, such as the preemptive reporting application described above.
1o The network service provider need only concern itself with detecting congestion and detecting usage.
The present invention has been described with respect to the operational condition of the network itself. Rather than focusing on a condition such as network congestion, the idea can be generalized to encompass "service" or "application-level"
congestion, i.e.
congestion as an attribute of a service or application as opposed to a network. The incentive information can be generated as a function of some condition of a service/application provided by the network operator to the customer.
For example, it is known that the load on e-mail servers increases near Christmastime, as users send each other holiday greetings laden with graphics and audio.
2o See "Graphic-Laden Holiday Greetings Clog Servers at AT&T WorldNet", Wall Street Journal, December 10, 1998. As the added multimedia makes the messages much larger than regular text e-mail, the mail servers become overloaded resulting in significant slowdowns in the delivery of incoming and outgoing mail. The present invention can be utilized to detect the increasing load on the mail server and to notify and charge customers submitting large messages during such periods of service congestion.
Similarly, receiving and storing certain high-volume newsgroups can be a considerable burden on a network operator trying to maintain a news server for its customers. Users who access such newsgroups by reading or posting to them could be notified and billed accordingly.
Service congestion detection can be accomplished by modifying the mail protocol or the netnews protocol. Where a change in the protocol is undesirable, fake congestion indicators can be generated. While artificially dropping packets in order to signal service-level congestion would be counterproductive, an ECN congestion bit can be set to notify the user of the service-level congestion. This scheme will work for any application where most of the data is sent from the server to the user, since it will then have a minimal effect on actual transmission speeds.
Although the embodiments of the present invention are described with respect to the Internet, it would be easily recognized by one of ordinary skill in the art that the to present invention is applicable to packet-switched networks in general. The foregoing Detailed Description is to be understood as being in every respect illustrative and exemplary, but not restrictive, and the scope of the invention disclosed herein is not to be determined from the Detailed Description, but rather from the claims as interpreted according to the full breadth permitted by the patent laws. It is to be understood that the 15 embodiments shown and described herein are only illustrative of the principles of the present invention and that various modifications may be implemented by those skilled in the art without departing from the scope and spirit of the invention.
NETWORKS
Field of the Invention The present invention relates to communication networks and, more particularly, to methods and systems for reducing congestion in a packet-switched or hybrid network.
background of the Invention With the explosive growth of today's information superhighway have come the inevitable traffic jams. Congestion is a serious problem today on the Internet, a worldwide system of computer networks using packet-switching technology to transfer 1o messages between computers. Packet-switching protocols such as the Transmission Control Protocol/Internet Protocol (TCP/IP) divide messavges into packets which travel along a path in the network that can be varied as conditions in the network change.
Specifically, TCP/IP, as currently implemented in the Internet, routes packets independently of each other, utilizing its best efforts without any specific concept of a 15 "connection". Accordingly, in the Internet, there is little :notion of "quality of service", no notion of guaranteed throughput, and no notion of bounded transmission delay.
Current implementations of TCP/IP rely on packet loss as an indicator of congestion in the network. As the network experiences congestion, data flowing through a network router becomes bottlenecked in a queue until tl~e queue overflows and packets 2o are lost. Load reduction is accomplished by utilizing a well-known "congestion avoidance" algorithm first described by Van Jacobson in 1988. See "Congestion Avoidance and Control", V. Jacobson, ACM SIGCOMM-88, August 1988, p. 314-29;
"TCP Slow Start, Congestion Avoidance, Fast Retransmit:, and Fast Recovery Algorithms", W. Stevens, RFC 2581 (revision of RFC 2001). In what is coined a "slow 25 start", a TCP source begins inserting packets into the network by starting with a minimal congestion window, allowing at most one unacknowledged packet in the network.
Each time an acknowledgement (ACK) is received, the congestion window is enlarged exponentially until a first threshold is reached or until a packet is dropped.
If the first threshold is reached, the TCP source continues to enlarge the congestion window linearly until either a second threshold is reached - or until a packet is dropped.
Upon the timeout of a retransmit timer (thereby indicating a dropped packet), the TCP source reduces the transmission rate and "backs ofd' to its minimal window, with the goal of allowing the network to reach some form of equilibrium.
As traffic on the Internet increases and more applications are run which are 1o sensitive to the delay caused by dropped packets (e.g. stre;aming audio and video), proposals have emerged to add some form of explicit congestion notification (ECN) to TCP. See "TCP and Explicit Congestion Notification", ACM Computer Communication Review, V. 24 N. 5, ~ctober 1994, p. 10-23; "A Proposal to add Explicit Congestion Notification (ECN) to IP", I~.I~. Ramakrishnan and Sally Floyd, RFC 2481. For example, 15 a network router with a queue nearing an overflow, rather than merely waiting for a packet to drop, can transmit a signal (in the form of a special bit in the packet's header) to indicate the presence of network congestion. The receiver's acknowledgment packet passes the notification on to the sender, which in turn slows down its transmission rate.
These methods of controlling congestion by signaling for a reduction in 2o transmission rate, however, do not address the root problem - namely, insufficient transmission capacity to support the explosive growth in the number of users demanding access to the Internet at the same time. Short of increasing the capacity of the network, these methods of addressing congestion can make performance degradation more gradual, but they cannot prevent it altogether. It would be preferable to establish a system that 25 reduced congestion in the network by affecting the network usage habits of the people accessing the Internet, and thereby directly addressing the problem of overrunning the capacity of the network.
Summary of the Invention The present invention permits a network service provider to detect an operational condition - such as congestion - in a packet-switched network and to alleviate such congestion by providing customer incentives to avoid use of the network. The detection mechanism triggers an incentive such as the modification of the user's access charges and the customer can be immediately notified of either the occurrence of the congestion or of information regarding the incentive. Usage of the network during congested periods can be deterred by imposing additional access charges during such periods -similarly, customers can be given a discount to encourage usage during periods of low congestion.
1o An incentive schedule can be tailored to dynamically change the usage patterns of the customers of the network to accommodate the operational conditions in the network. The present invention has application in the Internet, where a detection/notification mechanism, for example, can be implemented in a network node such as a router or in a network endpoint such as a client machine or a proxy or mail server.
In accordance with one aspect of the present invention there is provided a method, for use by a network access service provider to alleviate an undesired operational condition in the network, said method comprising the steps of: detecting the undesired operational condition in the network; detecting current usage of the network by a customer; and providing a financial incentive to the customer to cease current use of the 2o network and attempt use at another time when the undesired operational condition is not present.
These and other advantages of the invention will be apparent to those of ordinary skill in the art by reference to the following detailed description and the accompanying drawings.
Brief Description of the Drawings Fig. 1 is a block diagram of a packet-switched network illustrating an embodiment of the present invention.
Fig. 2 is a diagram showing the movement of illu:~trative packets in a packet-switched network between a sender and a receiver as a function of time.
Fig. 3 is a block diagram of a packet-switched network and a proxy server illustrating another embodiment of the present invention.
Fig. 4 is a block diagram of a packet-switched network illustrating another embodiment of the present invention.
Fig. 5 is a diagram of a user display with a popup window in the upper right hand corner of the display showing information regarding congestion in a packet-switched network.
Detailed Description The present invention is illustrated with reference to Fig. 1 which shows a packet-switched network 100 having numerous packet-switching nodes 110 to 119 connecting endpoints 101 and 102. Endpoint 101 and nodes 110 to 115 are assumed to be under the control or supervision of a network service provider; the remainder of the network is 1o assumed to be controlled or maintained by other providers or entities. The network service provider provides access for users to the network for an access charge. For example, where the network 100 is the Internet, an Internet Service Provider ("ISP'~ such as AT&T WorldNetT'" provides access to the Internet for its customers. In the case of the Internet, 110 to 115 represent routers and endpoints, 101 and 102 can be client machines, 15 servers, proxy servers, mail servers, news servers, etc. Fig. 1, of course, is a simplification as a typical communication network would encompass other network elements that would be apparent to one of ordinary skill in the art.
Furthermore, although the discussion below focuses on service providers, one of ordinary skill would easily recognize that the present invention applies equally to other network entities such as, for 2o example, corporate networks that utilize charge-back schemes.
In accordance with the present invention, operational conditions such as congestion in the network are detected in a network node 110 to 115 and/or at an endpoint 101 in the network under the control of the network service provider.
Upon the detection of the condition, whether at a router or an endpoint, the present invention 25 generates incentive information, such as billing records reflecting a reduction or an increase in the access charge paid by a particular user. This information can be relayed to a billing server or some other billing apparatus for processing. A
notification mechanism permits the user to receive notice of the incentive, either by notifying the user of the presence of the operational condition (e.g. congestion) or of the incentive information (e.g. the modified access charge).
The specific detection mechanism will depend on the particular operational condition in the network sought to be detected as well as the protocols that can be used in the network to signal the condition. For example, each node 110-115 is customarily equipped in a packet-switched network with a large number of buffers for storing packets in anticipation of routing or awaiting availability of an output link. With regard to packet congestion, such symptoms develop first in the node's buffers or queues, as the buffers become filled and unavailable to store incoming packets. Thus, a router knows that the 1o Internet is getting congested because its buffer queue for some output link is too long or is getting too long. Some routers today utilize a mechanism called Random Early Detection ("RED") which signals the presence of congestion as it develops by dropping packets when the average queue length exceeds some threshold - rather than waiting for the queue to actually overflow. See RFC 2309, which is incorporated by reference 15 herein.
Where endpoint 101 is a sender of packets across the network, see Fig. 2, it will also be aware of the congestion developing in the router's buffers. Endpoint 141, using TCP/IP, expects to receive an ACK 202 after transmitting a packet 201 through the network. Failure to receive an ACK signifies that the packet has been dropped by some 2o router between it and the destination endpoint 102. Where, however, endpoint 101 is a receiver of packets (and endpoint 102 is, accordingly, the sender), the situation is a bit more subtle. The TCP process layer at endpoint 101, as it receives the packets 201, 203, etc. sent by 102, knows the order in which to reassemble the packets based on a sequence number in the received packets' headers. Endpoint 101, thus, expects to receive the 25 packets in a certain order and can infer the dropping of a packet by looking for "holes" in the packets' sequence numbers. An out-of sequence packet, especially if there is a significant delay before the hole is filled, in general indicates that the expected packet has been dropped due to congestion. This method does not guarantee absolute detection of every dropped packet since, for example, packet loss will be invisible to the receiver if the trailing packet/packets in a sequence are dropped. Nevertheless, the method statistically provides good detection of dropped packets, especially for long transmissions (which is the situation a network service provider would be the most concerned about).
Alternatively, the endpoint could use duplicate packets as an indicator, although the method would not be expected to be as good as a method based on detecting a hole and a timeout.
The situation is simplified if the network has ECN capabilities. In that case, where the muter experiences congestion in its buffers (whether by a buffer overflow or by RED), it sets a "Congestion Experienced" ("CE") bit in the packet header of packets 1o from ECN-capable transport protocols. See RFC 2481. The receiver of the packet detects the CE bit and sets a "ECN-Echo" flag in the header of a subsequent ACK packet sent back to the sender. Endpoints 101 and 102 are thus quickly notified of the congestion condition in a router and can react accordingly.
The above detection mechanisms have been described with respect to the Internet 15 protocol suite although, as noted above, the present invention works with applications and protocols other than reliable data transfer over TCP (as well as non-TCP/IP networks such as Ethernet, hybrid networks, etc.). For example, the instant methods of congestion-based incentives work with multicast communication as well as unicast communication.
Consider a multicast audio application that runs over RTP/(TDP (the Real Time Protocol 2o over the User Datagram Protocol). In multicast audio, a sender transmits a stream of packets containing audio samples to multiple receivers. These applications do not require that every audio sample be reliably delivered, but they do require some reliability in order to maintain acceptable audio quality. RTP in particular uses sequence numbers to order packets and detect losses at each receiver. A lost RTP packet signals congestion just as a 25 lost TCP packet does. RTP packets can also arrive at a receiver, in principle, containing an ECN (Explicit Congestion Notification) signal placed there by the network.
Furthermore, RTP receivers send periodic reports back to RTP senders. Senders use these reports to monitor communication quality and possibly adapt their behavior when there are problems. RTP reports include loss information, just as TCP
acknowledgements do. Report packets can also carry ECN-Echo signals back to the sender. Therefore, both implicit and explicit congestion signals can be used in the context of the present invention as akeady described above.
To see how the present invention can be implemented in a real-world setting such as the Internet, it is necessary to understand the typical operating environment.
About 70 to 75% of the traffic in the Internet today utilizes the Hypertext Transfer Protocol (HTTP), i.e., Web page retrieval. Furthermore, most congestion in the case of HTTP will occur on user-bound packets, since that is the direction of most Web traffic.
Some users connect directly to the Web server of interest; others go through what are 1o known as proxy servers. Often operated by network service providers, a proxy server acts as an intenmediary between a user terminal and the Internet to provide caching services. By caching frequently-requested pages, a proxy server can reduce the bandwidth necessary for the provider's own connection to the outside world.
The present invention, then, should handle the case of both direct and proxy connections to the 15 outside world, especially (but not exclusively) for Web traffic.
First, consider the case of a user of a proxy server. With reference to Fig.
3, proxy server 301 receives a request from the user 300 for an Internet service (e.g. a Web page request from server 302) which the proxy server forwards through the Internet 305 using its own IP address - assuming the request passes filtering requirements and cannot 2o be satisfied by the internal cache. When the proxy server receives the requested page, it relates it to the original request and forwards it back to the relevant user.
A proxy server, as a receiver of packets from outside Web servers, can use the present invention to detect congestion in the network and attribute it to the customer who requested the specific Web page. It is also a sender on the provider's own network; this, too, can be noted 25 appropriately. The proxy server is thus perfectly positioned to detect and charge for requests over congested networks. T'he same analysis holds for operator-provided e-mail and, to a certain extent, news servers maintained by the operator. When sending or receiving mail on behalf of a given customer, operational conditions internal and external to the network can be detected, noted, and billed accordingly. (It should be noted, however, that there could be philosophical problems with regard to the application of the present invention to the receipt of unsolicited mail during periods of congestion; see the discussion below on service-level congestion.) Suppose, though, that the user does not utilize a proxy server and is connecting directly to the Internet, whether for Web-browsing or something else. See Fig.
4. In this case, the end systems that detect the congestion most easily are the customer's machine 401 and some endpoint 402 not under control of the network operator. While the former is ideally placed to notify the user, it clearly cannot be trusted to generate charge records for purposes of billing. Under certain circumstances, it may be possible to gather data to reliably, such as when the customer utilizes a tamper-resistant modem supplied by the network operator. Otherwise, the routers 410 to 41 S in Fig. 4 operated by the network service provider should be used to detect congestion and gather the data necessary to generate the appropriate billing records.
The only case not covered, then, is when congestion occurs, but on a router not 15 under the control of the local network operator, i.e. routers 416 to 419.
Arguably, this is not a matter of local concern, since the local network operator is not paying for the bandwidth. There is a situation of interest, however - namely, when the router detecting the congestion is at the other side of a comparatively slow link between the local network operator and an upstream service provider. There are two possible ways to overcome the 2o problem: First, the local operator can provide the routers for both ends of the slow link.
At the upstream end, a fast link can connect to the local provider's router;
this, then, reduces to the previous case. Alternatively, by contractual arrangement the upstream provider can detect and record congestion on behalf of the local operator.
This situation can be generalized. Network service providers can detect congestion attributable to other 25 customers of other network operators, and notify and bill them appropriately. A special-purpose "congestion indication protocol" can be utilized to pass the information between network operators. Accordingly, all users of an upstream provider can be charged for the congestion that they cause, rather than trying to attribute the problem to individual users.
Clearly, notifying the users of the incentive is important: if users do not realize there is a pmblem in the network (and a surcharge/reward), they will not modify their behavior at appropriate times. Direct notification from the network provider is straightforward where the present invention is implemented as a user process running on a direct endpoint capable of detecting congestion. With other configurations, other mechanisms for notifying the customer can be utilized. For example, when the user is in contact with only some local server, service-specific mechanisms can be utilized. When the customer is accessing Web pages through a proxy, a Java or Javascript applet can be sent to the user in the first Web page retrieved which, in turn, displays to the user the relevant notifications regarding congestion and incentive information such as the effect on access charges. Similarly, the mail retrieval protocol could be modified to send appropriate information to the customer when accessing a mail server to check the user's e-mail (or create a new e-mail).
The notification can be in the form of a window or screen "popup" on the user's display. A small window could be displayed on the user terminal indicating the level of congestion in the network in some visually intuitive and appealing form such as a speedometer. See Fig. 5. Moreover, other forms of notification can come from pre-emptive detection mechanisms. In commonly assigned U.S. Patent No. 5,870,557, "Method for Determining and Reporting a Level of Network Activity on a 2o Communications Network Using a Routing Analyzer", the disclosure of which is incorporated by reference herein, a method is disclosed for periodically analyzing the congestion along routes from an access provider's entry point to the Internet to a set of desired Web sites. Round trip times and packet loss information collected from use of the "traceroute" command are utilized to identify congested links. The transit characteristics are compiled and analyzed to provide a user with useful information about congestion along routes to the Web sites of interest and in order to warn users to avoid traffic to some Web sites at congested periods of time. Such preemptive mechanisms can be utilized with the present invention to provide a customer with prior information on the operational condition of the network before incurring any charges for usage of a congested link. Accordingly, a customer can be prompted before deciding to access the network in a manner that will generate any incentive information such as a modification to a customer billing record.
Moreover, it should be apparent from the above example that the notification mechanism need not be coupled to the network's detection and incentive mechanisms.
Self notification mechanisms can be utilized by the user. As long as users have received some notice of the general contours of the incentive, perhaps at registration, they can utilize whatever local detection/notification application scheme on their own computer terminal they choose to, such as the preemptive reporting application described above.
1o The network service provider need only concern itself with detecting congestion and detecting usage.
The present invention has been described with respect to the operational condition of the network itself. Rather than focusing on a condition such as network congestion, the idea can be generalized to encompass "service" or "application-level"
congestion, i.e.
congestion as an attribute of a service or application as opposed to a network. The incentive information can be generated as a function of some condition of a service/application provided by the network operator to the customer.
For example, it is known that the load on e-mail servers increases near Christmastime, as users send each other holiday greetings laden with graphics and audio.
2o See "Graphic-Laden Holiday Greetings Clog Servers at AT&T WorldNet", Wall Street Journal, December 10, 1998. As the added multimedia makes the messages much larger than regular text e-mail, the mail servers become overloaded resulting in significant slowdowns in the delivery of incoming and outgoing mail. The present invention can be utilized to detect the increasing load on the mail server and to notify and charge customers submitting large messages during such periods of service congestion.
Similarly, receiving and storing certain high-volume newsgroups can be a considerable burden on a network operator trying to maintain a news server for its customers. Users who access such newsgroups by reading or posting to them could be notified and billed accordingly.
Service congestion detection can be accomplished by modifying the mail protocol or the netnews protocol. Where a change in the protocol is undesirable, fake congestion indicators can be generated. While artificially dropping packets in order to signal service-level congestion would be counterproductive, an ECN congestion bit can be set to notify the user of the service-level congestion. This scheme will work for any application where most of the data is sent from the server to the user, since it will then have a minimal effect on actual transmission speeds.
Although the embodiments of the present invention are described with respect to the Internet, it would be easily recognized by one of ordinary skill in the art that the to present invention is applicable to packet-switched networks in general. The foregoing Detailed Description is to be understood as being in every respect illustrative and exemplary, but not restrictive, and the scope of the invention disclosed herein is not to be determined from the Detailed Description, but rather from the claims as interpreted according to the full breadth permitted by the patent laws. It is to be understood that the 15 embodiments shown and described herein are only illustrative of the principles of the present invention and that various modifications may be implemented by those skilled in the art without departing from the scope and spirit of the invention.
Claims (22)
1. A method, for use by a network access service provider to alleviate an undesired operational condition in the network, said method comprising the steps of:
detecting the undesired operational condition in the network;
detecting current usage of the network by a customer; and providing a financial incentive to the customer to cease current use of the network and attempt use at another time when the undesired operational condition is not present.
detecting the undesired operational condition in the network;
detecting current usage of the network by a customer; and providing a financial incentive to the customer to cease current use of the network and attempt use at another time when the undesired operational condition is not present.
2. The method of claim 1 wherein the financial incentive is a modification in an access charge collected by the network access service provider from the customer.
3. The method of claim 1 wherein the undesired operational condition is congestion in the network.
4. The method of claim 3 wherein congestion is detected by determining that a packet has been dropped in the network.
5. The method of claim 3 wherein congestion is detected by determining that a network endpoint has failed to receive an acknowledgement in reply to sending a packet.
6. The method of claim 3 wherein congestion is detected by determining that a network endpoint has received a packet out of sequence.
7. The method of claim 3 wherein congestion is detected by determining the size of a buffer queue of a network node.
8. The method of claim 3 wherein congestion is detected by determining that a packet has been marked by explicit congestion notification.
9. The method of claim 1 wherein the step of detecting the operational condition in the network occurs in a network endpoint.
10. The method of claim 9 wherein the network endpoint is a server.
11. The method of claim 9 wherein the network endpoint is a proxy server.
12. The method of claim 9 wherein the network endpoint is a mail server.
13. The method of claim 9 wherein the network endpoint is a news server.
14. The method of claim 9 wherein the network endpoint is a tamper-resistant modem.
15. The method of claim 1 wherein the step of detecting the unwanted operational condition in the network occurs in a network node.
16. The method of claim 15 wherein the network node is a router.
17. The method of claim 1 further comprising the step of:
providing an access charge discount incentive notification to the customer to use the network at another time when the undesired operational condition is not present.
providing an access charge discount incentive notification to the customer to use the network at another time when the undesired operational condition is not present.
18. The method of claim 1 wherein the network is a packet-switched network.
19. The method of claim 1 wherein the network is a hybrid network.
20. The method of claim 18 wherein the network uses the Internet protocol suite.
21. The method of claim 18 wherein the network uses a multicast communication protocol.
22. The method of claim 18 wherein the network is the Internet.
Applications Claiming Priority (2)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
US29202499A | 1999-04-16 | 1999-04-16 | |
US09/292,024 | 1999-04-16 |
Publications (2)
Publication Number | Publication Date |
---|---|
CA2301435A1 CA2301435A1 (en) | 2000-10-16 |
CA2301435C true CA2301435C (en) | 2006-10-10 |
Family
ID=23122856
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CA002301435A Expired - Fee Related CA2301435C (en) | 1999-04-16 | 2000-03-14 | Method for reducing congestion in packet-switched networks |
Country Status (5)
Country | Link |
---|---|
US (2) | US7227843B2 (en) |
EP (1) | EP1045546B1 (en) |
CA (1) | CA2301435C (en) |
DE (1) | DE60018829T2 (en) |
MX (1) | MXPA00003614A (en) |
Families Citing this family (39)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20130343205A1 (en) * | 1999-04-01 | 2013-12-26 | Callwave Communications, Llc | Methods and apparatus for providing expanded telecommunications service |
US7822188B1 (en) | 1999-04-01 | 2010-10-26 | Callwave, Inc. | Methods and apparatus for providing expanded telecommunications service |
US6668051B1 (en) * | 2000-03-28 | 2003-12-23 | Sevis Systems, Inc. | Intelligent communications point platform |
DE10019000B4 (en) * | 2000-04-17 | 2004-11-18 | Siemens Ag | Procedure for recording usage fees |
US6778961B2 (en) * | 2000-05-17 | 2004-08-17 | Wconect, Llc | Method and system for delivering text-to-speech in a real time telephony environment |
US6898182B1 (en) * | 2000-07-21 | 2005-05-24 | Arris International, Inc | Congestion control in a network device having a buffer circuit |
US7562147B1 (en) * | 2000-10-02 | 2009-07-14 | Microsoft Corporation | Bi-directional HTTP-based reliable messaging protocol and system utilizing same |
US7720951B2 (en) * | 2001-05-10 | 2010-05-18 | International Business Machines Corporation | Method, system, and product for alleviating router congestion |
US7050393B2 (en) * | 2001-05-10 | 2006-05-23 | International Business Machines Corporation | Method, system, and product for alleviating router congestion |
US7054940B2 (en) | 2002-01-25 | 2006-05-30 | Thomson Licensing | Adaptive cost of service for communication network based on level of network congestion |
US7571317B1 (en) * | 2002-09-11 | 2009-08-04 | Cisco Technology, Inc. | Providing user notification signals in phones that use encryption |
WO2004068800A1 (en) * | 2003-01-28 | 2004-08-12 | Telefonaktiebolaget Lm Ericsson (Publ) | Method and device for congestion notification in packet networks indicating several different congestion causes |
US7463652B2 (en) * | 2003-06-21 | 2008-12-09 | Avaya, Inc. | System and method for notification of internet users about faults detected on an IP network |
US20050060423A1 (en) * | 2003-09-15 | 2005-03-17 | Sachin Garg | Congestion management in telecommunications networks |
GB0321570D0 (en) | 2003-09-15 | 2003-10-15 | British Telecomm | Inter-domain congestion charging |
JP4415317B2 (en) * | 2004-02-23 | 2010-02-17 | セイコーエプソン株式会社 | Information equipment and control method thereof |
KR20070085208A (en) * | 2004-08-17 | 2007-08-27 | 캘리포니아 인스티튜트 오브 테크놀로지 | Method and apparatus for network congestion control using queue control and one-way delay measurements |
US7564796B2 (en) * | 2004-09-30 | 2009-07-21 | Hewlett-Packard Development Company, L.P. | Method and system for managing a network slowdown |
US7801127B2 (en) * | 2004-10-25 | 2010-09-21 | Ineoquest Technologies, Inc. | System and method for creating a sequence number field for streaming media in a packet-based networks utilizing internet protocol |
US7733770B2 (en) * | 2004-11-15 | 2010-06-08 | Intel Corporation | Congestion control in a network |
US7515543B2 (en) * | 2004-12-21 | 2009-04-07 | Alcatel-Lucent Usa Inc. | Packet reorder resolution in a load-balanced network architecture |
US8670309B2 (en) * | 2005-09-30 | 2014-03-11 | Alcatel Lucent | Method and apparatus for preventing activation of a congestion control process |
US20070230369A1 (en) * | 2006-03-31 | 2007-10-04 | Mcalpine Gary L | Route selection in a network |
JP4280272B2 (en) * | 2006-05-31 | 2009-06-17 | 株式会社東芝 | Information processing device |
US8018933B2 (en) | 2007-06-27 | 2011-09-13 | Microsoft Corporation | Reliable multicast with automatic session startup and client backfil support |
US8612617B2 (en) * | 2007-06-28 | 2013-12-17 | Microsoft Corporation | Reliable multicast transport protocol |
US8683065B2 (en) * | 2007-06-29 | 2014-03-25 | Microsoft Corporation | Multicast content provider |
US7882240B2 (en) * | 2007-07-03 | 2011-02-01 | Microsoft Corporation | Disconnecting selected participant in multicast session |
CN101534519B (en) * | 2008-03-14 | 2014-03-12 | 摩托罗拉移动公司 | Method for displaying package switching congestion state of wireless communication network |
US8565249B2 (en) * | 2009-02-10 | 2013-10-22 | Telefonaktiebolaget L M Ericsson (Publ) | Queue management system and methods |
US8873717B2 (en) | 2010-11-23 | 2014-10-28 | Centurylink Intellectual Property Llc | Emergency alert signaling |
EP2813007A4 (en) * | 2012-02-06 | 2015-10-28 | Intel Corp | Signaling for configuration of downlink coordinated multipoint communications |
US9013996B2 (en) | 2012-05-16 | 2015-04-21 | Centurylink Intellectual Property Llc | Customer configuration of broadband services |
US20140179265A1 (en) * | 2012-12-21 | 2014-06-26 | Glenn R. Bruns | Method For Offering Wireless Data At A Discounted Rate |
US9444741B2 (en) * | 2013-03-11 | 2016-09-13 | Broadcom Corporation | Facilitating network flows |
US9647950B2 (en) * | 2015-05-11 | 2017-05-09 | Ebay Inc. | System and method of site traffic control |
US9843517B2 (en) | 2015-05-14 | 2017-12-12 | Qualcomm Incorporated | Dynamically adjusting network services stratum parameters based on access and/or connectivity stratum utilization and/or congestion information |
DK3320652T3 (en) * | 2015-07-06 | 2019-11-04 | Ericsson Telefon Ab L M | PROCEDURE FOR CUSTOMER NETWORK COOPERATION BASED ON NON-ECONOMIC INCENTIVE |
KR102358344B1 (en) * | 2016-11-03 | 2022-02-07 | 프라운호퍼 게젤샤프트 쭈르 푀르데룽 데어 안겐반텐 포르슝 에. 베. | Network-based download/streaming concept |
Family Cites Families (37)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US4486626A (en) * | 1982-06-29 | 1984-12-04 | At&T Technologies, Inc. | Method of and system for limiting access to a group of telephone trunks |
US5029164A (en) * | 1990-04-13 | 1991-07-02 | Digital Equipment Corporation | Congestion avoidance in high-speed network carrying bursty traffic |
US5515425A (en) * | 1993-01-19 | 1996-05-07 | At&T Corp. | Telecommunications system with active database |
US5561703A (en) * | 1994-07-06 | 1996-10-01 | Rolm Company | System and method for integration of a paging server into a private branch exchange environment |
EP0734185A2 (en) * | 1995-03-23 | 1996-09-25 | Hughes Electronics | DTMF tone transmission in a voice communication system |
US6304574B1 (en) * | 1995-06-07 | 2001-10-16 | 3Com Corporation | Distributed processing of high level protocols, in a network access server |
US5862203A (en) * | 1995-07-21 | 1999-01-19 | Call Manage | Telecommunications call management system |
US6542502B1 (en) * | 1996-01-26 | 2003-04-01 | International Business Machines Corporation | Multicasting using a wormhole routing switching element |
US5930474A (en) * | 1996-01-31 | 1999-07-27 | Z Land Llc | Internet organizer for accessing geographically and topically based information |
US6353596B1 (en) * | 1996-04-12 | 2002-03-05 | Lucent Technologies Inc. | System and method for multipoint-to-multipoint multicasting |
US6442172B1 (en) * | 1996-07-11 | 2002-08-27 | Alcatel Internetworking, Inc. | Input buffering and queue status-based output control for a digital traffic switch |
US5870557A (en) * | 1996-07-15 | 1999-02-09 | At&T Corp | Method for determining and reporting a level of network activity on a communications network using a routing analyzer and advisor |
US5754786A (en) * | 1996-09-13 | 1998-05-19 | At&T Corp | System and method for integrated overload control and message distribution |
US5974308A (en) * | 1996-11-13 | 1999-10-26 | Telefonaktiebolaget Lm Ericsson | Selective broadcasting of charge rates |
US6188690B1 (en) * | 1996-12-12 | 2001-02-13 | Pmc-Sierra, Inc. | Method and apparatus for high speed, scalable communication system |
US6038593A (en) * | 1996-12-30 | 2000-03-14 | Intel Corporation | Remote application control for low bandwidth application sharing |
US6085892A (en) * | 1997-02-07 | 2000-07-11 | Quantum Conveyor Systems, Inc. | High speed sorting/diverting apparatus, an apparatus controller and systems using same |
JP2865139B2 (en) * | 1997-04-18 | 1999-03-08 | 日本電気株式会社 | ATM cell buffer circuit and arbitrary priority allocation method in ATM exchange |
JPH113307A (en) * | 1997-06-13 | 1999-01-06 | Canon Inc | Information processor and its method |
US6134589A (en) * | 1997-06-16 | 2000-10-17 | Telefonaktiebolaget Lm Ericsson | Dynamic quality control network routing |
US6408003B1 (en) * | 1997-06-17 | 2002-06-18 | Qualcomm Incorporated | Method and apparatus for resolving ambiguity in reception of multiple retransmitted frames |
US6757268B1 (en) * | 1997-07-21 | 2004-06-29 | Winstar Corporation | Metropolitan wide area network |
US6295294B1 (en) * | 1997-08-07 | 2001-09-25 | At&T Corp. | Technique for limiting network congestion |
US6385644B1 (en) * | 1997-09-26 | 2002-05-07 | Mci Worldcom, Inc. | Multi-threaded web based user inbox for report management |
US6389005B1 (en) * | 1997-12-01 | 2002-05-14 | Nortel Networks Limited | Automatic backup trunking for voice over the internet |
US6438101B1 (en) * | 1997-12-23 | 2002-08-20 | At&T Corp. | Method and apparatus for managing congestion within an internetwork using window adaptation |
US6233237B1 (en) * | 1998-02-02 | 2001-05-15 | 3Com Corporation | Method and protocol for connecting data calls using R2 signaling |
US6170061B1 (en) * | 1998-02-04 | 2001-01-02 | 3Com Corporation | Method and system for secure cable modem registration |
US6959318B1 (en) * | 1998-03-06 | 2005-10-25 | Intel Corporation | Method of proxy-assisted predictive pre-fetching with transcoding |
US6275471B1 (en) * | 1998-05-12 | 2001-08-14 | Panasonic Technologies, Inc. | Method for reliable real-time multimedia streaming |
US6526022B1 (en) * | 1998-06-30 | 2003-02-25 | Sun Microsystems | Detecting congestion by comparing successive loss of packets in windows to provide congestion control in reliable multicast protocol |
US6505253B1 (en) * | 1998-06-30 | 2003-01-07 | Sun Microsystems | Multiple ACK windows providing congestion control in reliable multicast protocol |
US6286098B1 (en) * | 1998-08-28 | 2001-09-04 | Sap Aktiengesellschaft | System and method for encrypting audit information in network applications |
US6160793A (en) * | 1998-10-13 | 2000-12-12 | Nokia Telecommunications, Oy | ECN-based approach for congestion management in hybrid IP-ATM networks |
US6522875B1 (en) * | 1998-11-17 | 2003-02-18 | Eric Morgan Dowling | Geographical web browser, methods, apparatus and systems |
US6370376B1 (en) * | 1999-03-18 | 2002-04-09 | David J. Sheath | Computer cellular communication system |
WO2001080515A2 (en) * | 2000-04-17 | 2001-10-25 | Circadence Corporation | System and method for data prioritization |
-
2000
- 2000-03-14 CA CA002301435A patent/CA2301435C/en not_active Expired - Fee Related
- 2000-04-05 EP EP00107082A patent/EP1045546B1/en not_active Expired - Lifetime
- 2000-04-05 DE DE60018829T patent/DE60018829T2/en not_active Expired - Lifetime
- 2000-04-13 MX MXPA00003614A patent/MXPA00003614A/en unknown
-
2003
- 2003-02-24 US US10/372,624 patent/US7227843B2/en not_active Expired - Fee Related
-
2007
- 2007-04-25 US US11/789,584 patent/US20070258375A1/en not_active Abandoned
Also Published As
Publication number | Publication date |
---|---|
US20070258375A1 (en) | 2007-11-08 |
US7227843B2 (en) | 2007-06-05 |
EP1045546A2 (en) | 2000-10-18 |
CA2301435A1 (en) | 2000-10-16 |
US20030137938A1 (en) | 2003-07-24 |
MXPA00003614A (en) | 2002-03-08 |
EP1045546A3 (en) | 2003-08-20 |
DE60018829T2 (en) | 2006-02-09 |
DE60018829D1 (en) | 2005-04-28 |
EP1045546B1 (en) | 2005-03-23 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CA2301435C (en) | Method for reducing congestion in packet-switched networks | |
Floyd | Congestion control principles | |
US6341309B1 (en) | Firewall system for quality of service management | |
Eggert et al. | Effects of ensemble-TCP | |
US8130647B2 (en) | Method and apparatus for managing aggregate bandwidth at a server | |
US7333431B2 (en) | Method for accelerating TCP flows in communication networks with dynamic control | |
US8379515B1 (en) | TCP throughput control by imposing temporal delay | |
EP2772028B1 (en) | Control system, gateway and method for selectively delaying network data flows | |
Bonald | Comparison of TCP Reno and TCP Vegas via fluid approximation | |
US20070081543A1 (en) | Network utilization control apparatus and method of using | |
US10033653B2 (en) | Controlling a transmission control protocol congestion window size | |
US20100274893A1 (en) | Methods and apparatus for detecting and limiting focused server overload in a network | |
US7177277B2 (en) | Method for optimizing performance in wireless networks | |
US7545743B2 (en) | P2P traffic supporting router and P2P traffic information sharing system using the router | |
US6700876B1 (en) | Congestion monitoring and message flow control in a blocking network | |
Ndikumana et al. | Network-assisted congestion control for information centric networking | |
Sisalem et al. | The direct adjustment algorithm: A TCP-friendly adaptation scheme | |
EP1341350B1 (en) | A method for congestion detection for IP flows over a wireless network | |
US20100165838A1 (en) | Method and apparatus for improving data throughput in a network | |
Lee et al. | Explicit traffic shaping scheme in content centric networking | |
Koptchev et al. | Traffic and congestion control in a publish/subscribe network | |
Salih et al. | Management of Network Congestion using an Explicit Congestion Network (ECN) | |
Wu et al. | Improving ensemble-TCP performance on asymmetric networks | |
Pang et al. | A novel red-based hop-by-hop congestion control | |
Zheng | Adaptive Explicit Congestion Notification (AECN) for Heterogeneous Flows |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
EEER | Examination request | ||
MKLA | Lapsed |
Effective date: 20160314 |