Multiplex communications – Data flow congestion prevention or control – Control of data admission to the network
Reexamination Certificate
1998-10-07
2001-04-10
Olms, Douglas W. (Department: 2732)
Multiplex communications
Data flow congestion prevention or control
Control of data admission to the network
C370S412000, C709S234000
Reexamination Certificate
active
06215769
ABSTRACT:
BACKGROUND OF THE INVENTION
1. Field of the Invention
This invention relates in general to networks, and more particularly to an enhanced acknowledgment pacing device and method for TCP connections.
2. Description of Related Art
Today, an organization's computer network has become its circulatory system. Organizations have combined desktop work stations, servers, and hosts into Local Area Network (LAN) communities. These Local Area Networks have been connected to other Local Area Networks and to Wide Area Networks (WANs). It has become a necessity of day-to-day operation that pairs of systems must be able to communicate when they need to, without regard to where they may be located in the network.
During the early years of network computing, proprietary networking protocols were the standard. However, the development of the Open Systems Interconnection Reference Model introduced by the International Organization for Standardization (ISO) has led to an impressive degree of interworking, which generally allows end-user applications to work very well between systems in a network. Implementations are based on written standards that have been made available by volunteers from dozens of computer vendors, hardware component vendors and independent software companies.
During the last decade, LANs have been proliferating. This has created a recurring problem of how to minimize congestion and optimize throughput that must be solved by network managers. An early solution was to simply divide Local Area Networks into multiple smaller networks serving smaller populations. These segments were connected by bridges to form a single Local Area Network with traffic being segregated locally to each segment.
The evolution of new network types and Wide Area Networks created a need for routers. For example, the Internet is a set of networks connected by gateways, which are sometimes referred to as routers. Routers added filtering and firewalling capability to provide more control over broadcast domains, limit broadcast traffic and enhance security. A router is able to chose the best path through the network due to embedded intelligence. This added intelligence also allowed routers to build redundant paths to destinations when possible. Nevertheless, the added complexity of best path selection capability accorded by the embedded intelligence increased the port cost of routers and caused substantial latency overhead. Shared-media networks comprising distributed client/server data traffic, expanded user populations and more complex applications gave birth to new bandwidth bottlenecks. Such congestion produced unpredictable network response times, the inability to support the delay-sensitive applications and higher network failure rates.
Congestion control in modern networks is increasingly becoming an important issue. The explosive growth of Internet applications such as the World Wide Web (WWW) has pushed current technology to its limit, and it clear that faster transport and improved congestion control mechanisms are required. As a result, many equipment vendors and service providers are turning to advanced networking technology to provide adequate solutions to the complex quality of service (QoS) management issues involved. Examples include asynchronous transfer made (ATM) networks and emerging IP network services. Nevertheless, there is still the need to support a host of existing legacy IP protocols within these newer paradigms. In particular, the ubiquitous TCP transport-layer protocol has long been the workhorse transport protocol in IP networks, widely used by web-browsers, file/email transfer services, etc.
Transmission Control Protocol (TCP) is a part of the TCP/IP protocol family that has gained the position as one of the world's most important data communication protocols with the success of the Internet. TCP provides a reliable data connection between devices using TCP/IP protocols. TCP operates on top of IP that is used for packing the data to data packets, called datagrams, and for transmitting across the network.
The Internet Protocol (IP) is a network layer protocol that routes data across an Internet. The Internet Protocol was designed to accommodate the use of host and routers built by different vendors, encompass a growing variety of growing network types, enable the network to grow without interrupting servers, and support higher-layer of session and message-oriented services. The IP network layer allows integration of Local Area Network “islands”.
However, IP doesn't contain any flow control or retransmission mechanisms. That is why TCP is typically used on top of it. Especially, TCP uses acknowledgments for detecting lost data packets. TCP/IP networks arc nowadays probably the most important of all networks, and operate on top of several (physical) networks, such as the ATM networks mentioned above. These underlying networks may offer some information about the condition of network and traffic, which may be used to provide feedback regarding congestion.
To manage congestion, TCP uses a sliding window mechanism coupled with reactive congestion control to adjust the sender's window size. The protocol adjusts its transmission behavior contingent to returning acknowledgment (ACK) packets sent from the remote receiver's end.
A problem with TCP, however, is that its congestion control mechanism is relatively slow. Most TCP implementations use very coarse timers to measure timeouts, i.e., roughly 200-500 ms granularity. Further, most TCP implementations rely on ACK delays or packet drops to detect congestion. As a result, excessive source window reductions can result in large amounts of bandwidth being wasted as the TCP source is forced to restart its transmission window. Further, many studies have shown that TCP does not perform very well over ATM networks, especially for larger WAN-type propagation delays.
To combat the above shortcomings with TCP, it is necessary to minimize the chances of network congestion by somehow incorporating faster congestion indication mechanisms in the TCP feedback loop. However, to ensure compatibility with current versions and to expedite market acceptance, any such attempt must preclude changes to the actual TCP protocol or its implementation.
Along these lines, a variety of ACK pacing schemes have been proposed. These ACK pacing schemes basically modulate the spacing of TCP ACK packets to limit source emissions during periods of congestion. ACK pacing is well-suited at the boundary of high speed (sub)networks, such as ATM, gigabit IP (i.e., optical WDM), or satellite. In essence this technique performs TCP traffic shaping at the access nodes. Such methodologies are specifically beneficial for advanced ATM data services, i.e., underlying ABR flow control or per-connection queuing, where congestion tends to buildup at the periphery of the ATM network, i.e., in the access nodes. If the forward link is congested, as indicated via some congestion metric, ACK packets are appropriately delayed before being sent to the source.
Other authors have proposed modifying fields in the ACK packets themselves, i.e., receiver-window size, to improve performance. However, such schemes either require accurate round-trip delay measurements or cannot maintain tight buffer control. Furthermore, rewriting ACK packet fields will require expensive checksum recomputations.
Although ACK pacing is an effective way of controlling TCP source behaviors, many of the proposed schemes arc either too complex and/or overly sensitive to network parameter settings. Since studies have shown that TCP's throughput and fairness levels can be low in many high-speed network scenarios, it is necessary to devise efficient, practical schemes to enhance its performance. Although amending the protocol's functionality itself is also an option, this may not be a feasible alternative in the short-to-medium time frame. It is along these lines that the ACK pacing methods can provide significant benefits.
It can be seen that there is a need for a more robust, comprehensiv
Dixit Sudhir Sharan
Ghani Nasir
Altera Law Group LLC
Nokia Telecommunications, Inc.
Olms Douglas W.
Vanderpuye Ken
LandOfFree
Enhanced acknowledgment pacing device and method for TCP... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Enhanced acknowledgment pacing device and method for TCP..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Enhanced acknowledgment pacing device and method for TCP... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2468779