Improving TCP performance over wireless networks at the link layer

We present the transport unaware link improvement protocol (TULIP), which dramatically improves the performance of TCP over lossy wireless links, without competing with or modifying the transport- or network-layer protocols. TULIP is tailored for the half-duplex radio links available with today's commercial radios and provides a MAC acceleration feature applicable to collision-avoidance MAC protocols (e.g., IEEE 802.11) to improve throughput. TULIP's timers rely on a maximum propagation delay over the link, rather than performing a round-trip time estimate of the channel delay. The protocol does not require a base station and keeps no TCP state. TULIP is exceptionally robust when bit error rates are high; it maintains high goodput, i.e., only those packets which are in fact dropped on the wireless link are retransmitted and then only when necessary. The performance of TULIP is compared against the performance of the Snoop protocol (a TCP-aware approach) and TCP without link-level retransmission support. The results of simulation experiments using the actual code of the Snoop protocol show that TULIP achieves higher throughput, lower packet delay, and smaller delay variance. This revised version was published online in June 2006 with corrections to the Cover Date.     

Shielding TCP from last hop wireless losses

The pervasiveness of the transport control protocol (TCP) and the proliferation of wireless local area networks (WLAN) of the 802.11 type make the topic of TCP performance over last hop wireless networks very relevant. The Snoop protocol, a link layer solution introduced several years ago to improve the performance of TCP in this scenario, has been shown to neglect its benefits to the most widely used TCP version, TCP SACK. In this paper, we introduce the TCP SACK‐Aware Snoop protocol to address this problem. Our results indicate that the TCP SACK‐Aware Snoop protocol improves the performance of TCP SACK by around 30% compared with the original Snoop protocol and by about 8% in an environment where no TCP enhancing mechanism is in place. In addition, we introduce further modifications to the proposed protocol to make its advantages available to any TCP sender. We also show that the mechanism does not introduce unfairness among TCP sources and somewhat protects TCP against UDP traffic. Our results show important throughput improvements to all TCP versions and demonstrate that the TCP SACK‐Aware Snoop protocol shields TCP from last hop wireless losses providing throughtput values very close to the maximum possible. Copyright © 2006 John Wiley & Sons, Ltd.     

Performance study of link layer and MAC layer protocols to support TCP in 3G CDMA systems

Providing support for TCP with good quality link connection is a key issue for future wireless networks in which Internet access is going to be one of the most important data services. A number of schemes have been proposed in literature to improve the TCP performance over wireless links. In this paper, we study the performance of a particular combination of link layer protocol (e.g., radio link protocol or RLP) and MAC retransmissions to support the TCP connections over third generation (3G) wireless CDMA networks. We specifically investigate two metrics - the packet error rate and the delay provided by RLP and MAC retransmissions - both of which are important for TCP performance. For independent and identically distributed (i.i.d) error channels, we propose an analytical model for RLP performance with MAC retransmission. The segmentation of TCP/IP packets into smaller RLP frames, as well as the RLP buffering process, is modeled using a Markov chain. For correlated fading channels, we introduce an analytical metric called RLP retransmission efficiency. We show that: 1) the RLP frame size has significant impact on the overall 3G system performance, 2) MAC layer retransmissions significantly improve the TCP performance, and 3) the RLP retransmission scheme performs better in highly correlated channels, while other scheme performs better in low correlated channels. Simulation results also confirm these conclusions.     

A TCP acceleration algorithm for a wireless link using rate adaptation based on round-trip-time and virtual receiver window information

Fast transmission control protocol (TCP) was previously proposed for high capacity network environments with long delay, and "FAST TCP with Snoop" performs better than conventional TCP enhancements in mobile wireless network environments. However, FAST TCP has limitations when used over a dynamic mobile wireless link with a high frame error ratio (FER) and frequent delay changes due to the variable rate. We propose an enhanced TCP acceleration algorithm at the TCP sender side which efficiently adapts to the maximum transmission rate of a mobile wireless link using the round trip time (RTT) and virtual receiver window (RWND) information. The proposed algorithm provides superior performance over mobile wireless network environments.     

Vertical optimization of data transmission for mobile wireless terminals

A major problem for TCP connections over wireless links is that errors introduced by the wireless channel interfere with the TCP protocol, leading to reduced data rates and power wastage. Based on accurate simulations for the TCP and IEEE 802.11 MAC protocols, we discuss recipes to optimize transmission. It is argued that the best approach is to restrict modifications to the mobile device. While this requires separate solutions for the uplink and downlink, the results of optimization are then available when roaming into any WLAN obeying the relevant MAC protocol. Simulation results show that the combination of specific strategies with a vertical interaction between the protocol layers can lead to the required improvements, giving a promising approach to enhance the performance of wireless mobile terminals.     

Advocating a Remote Socket Architecture for Internet Access Using Wireless LANs

One challenge in the development of telecommunication networks is the seamless integration of wireless devices into the global Internet. Although it is well known that the Internet protocols were designed for heterogeneous networks an end-system with the usual Internet protocol stack will suffer an inefficient communication while connected via a wireless link. The protocol mechanisms of the transport layer can lead to poor performance in case of TCP and a high loss rate in case of UDP. In this paper we advocate a Remote Socket Architecture (ReSoA) which is a kind of proxy-oriented architecture for wireless Internet access in Wireless LAN environment. This approach allows the use of a thin protocol stack on the wireless end-system to save scarce resources and a tailored protocol for the wireless link without breaking the original TCP semantics. We show the suitability of ReSoA by comparing its performance with that of pure TCP and Berkeley Snoop through actual measurements in a test environment.     

Internet protocol performance over networks with wireless links

This article discusses the problems that arise when standard Internet protocols such as TCP are used over wireless links. We review wireless link characteristics with case studies drawn from commercial wireless LANs and cellular telephony systems. We discuss problems with Internet protocols when employed over these systems, such as degraded TCP performance when wireless errors are interpreted as congestion losses. We survey various proposed approaches to mitigating such problems and examine their applicability. Finally, we look at the future of wireless systems and the new challenges that they will create for Internet protocols, and state some goals for further protocol enhancement and evolution, pointing out the need for better protocol integration across layers     

TCP over wireless with link level error control: analysis anddesign methodology

This paper considers the problem of supporting TCP, the Internet data transport protocol, over a lossy wireless link whose quality varies over time. In order to prevent throughput degradation, it is necessary to “hide” the losses and the time variations of the wireless link from TCP. A number of solutions to this problem have been proposed in previous studies, but their performance was studied on a purely experimental basis. This paper presents an approximate analysis, validated by computer simulations, for TCP performance over wireless links. The analysis provides the basis for a systematic approach to supporting TCP over wireless links. The specific case of a Rayleigh-faded wireless link and automatic repeat request-based link-layer recovery is considered for the purpose of illustration. The numerical results presented for this case show that a simple solution, that of using an appropriately designed link-layer error-recovery scheme, prevents excessive deterioration of TCP throughput on wireless links     

Delayed duplicate acknowledgements: a TCP‐Unaware approach to improve performance of TCP over wireless

Since a TCP sender cannot distinguish between packet losses arising from transmission errors from those due to congestion, TCP tends to perform poorly on wireless links that are prone to transmission errors. Several techniques have previously been proposed to improve TCP performance over wireless links. Existing schemes typically require an intermediate node (typically, a base station) to be TCP‐aware. For instance, the Snoop scheme requires the base station to interpret TCP headers and take appropriate action to help improve TCP performance. This paper proposes an alternative TCP‐unaware technique that attempts to mimic the behavior of the Snoop protocol. Performance evaluation shows that the proposed Delayed Dupacks scheme performs quite well. Copyright © 2001 John Wiley & Sons, Ltd.     

Wireless video streaming with TCP and simultaneous MAC packet transmission (SMPT)

Video streaming is expected to account for a large portion of the traffic in future networks, including wireless networks. It is widely accepted that the user datagram protocol (UDP) is the preferred transport protocol for video streaming and that the transmission control protocol (TCP) is unsuitable for streaming. The widespread use of UDP, however, has a number of drawbacks, such as unfairness and possible congestion collapse, which are avoided by TCP. In this paper we investigate the use of TCP as the transport layer protocol for streaming video in a multi‐code CDMA cellular wireless system. Our approach is to stabilize the TCP throughput over the wireless links by employing a recently developed simultaneous MAC packet transmission (SMPT) approach at the link layer. We study the capacity, i.e. the number of customers per cell, and the quality of service for streaming video in the uplink direction. Our extensive simulations indicate that streaming over TCP in conjunction with SMPT gives good performance for video encoded in a closed loop, i.e. with rate control. We have also found that TCP is unsuitable (even in conjunction with SMPT) for streaming the more variable open‐loop encoded video. Copyright © 2004 John Wiley & Sons, Ltd.     

TCP Performance in IEEE 802.11-Based Ad Hoc Networks with Multiple Wireless Lossy Links

We propose a packet-level model to investigate the impact of channel error on the transmission control protocol (TCP) performance over IEEE-802.11-based multihop wireless networks. A Markov renewal approach is used to analyze the behavior of TCP Reno and TCP Impatient NewReno. Compared to previous work, our main contributions are listed as follows: 1) modeling multiple lossy links, 2) investigating the interactions among TCP, Internet Protocol (IP), and media access control (MAC) protocol layers, specifically the impact of 802.11 MAC protocol and dynamic source routing (DSR) protocol on TCP throughput performance, 3) considering the spatial reuse property of the wireless channel, the model takes into account the different proportions between the interference range and transmission range, and 4) adopting more accurate and realistic analysis to the fast recovery process and showing the dependency of throughput and the risk of experiencing successive fast retransmits and timeouts on the packet error probability. The analytical results are validated against simulation results by using GloMoSim. The results show that the impact of the channel error is reduced significantly due to the packet retransmissions on a per-hop basis and a small bandwidth delay product of ad hoc networks. The TCP throughput always deteriorates less than ~ 10 percent, with a packet error rate ranging from 0 to 0.1. Our model also provides a theoretical basis for designing an optimum long retry limit for IEEE 802.11 in ad hoc networks.     

A Smart TCP Acknowledgment Approach for Multihop Wireless Networks

Reliable data transfer is one of the most difficult tasks to be accomplished in multihop wireless networks. Traditional transport protocols like TCP face severe performance degradation over multihop networks given the noisy nature of wireless media as well as unstable connectivity conditions in place. The success of TCP in wired networks motivates its extension to wireless networks. A crucial challenge faced by TCP over these networks is how to operate smoothly with the 802.11 wireless MAC protocol which also implements a retransmission mechanism at link level in addition to short RTS/CTS control frames for avoiding collisions. These features render TCP acknowledgments (ACK) transmission quite costly. Data and ACK packets cause similar medium access overheads despite the much smaller size of the ACKs. In this paper, we further evaluate our dynamic adaptive strategy for reducing ACK-induced overhead and consequent collisions. Our approach resembles the sender side's congestion control. The receiver is self-adaptive by delaying more ACKs under nonconstrained channels and less otherwise. This improves not only throughput but also power consumption. Simulation evaluations exhibit significant improvement in several scenarios     

Improving reliable transport and handoff performance in cellular wireless networks

TCP is a reliable transport protocol tuned to perform well in traditional networks where congestion is the primary cause of packet loss. However, networks with wireless links and mobile hosts incur significant losses due to bit-errors and handoffs. This environment violates many of the assumptions made by TCP, causing degraded end-to-end performance. In this paper, we describe the additions and modifications to the standard Internet protocol stack (TCP/IP) to improve end-to-end reliable transport performance in mobile environments. The protocol changes are made to network-layer software at the base station and mobile host, and preserve the end-to-end semantics of TCP. One part of the modifications, called the snoop module, caches packets at the base station and performs local retransmissions across the wireless link to alleviate the problems caused by high bit-error rates. The second part is a routing protocol that enables low-latency handoff to occur with negligible data loss. We have implemented this new protocol stack on a wireless testbed. Our experiments show that this system is significantly more robust at dealing with unreliable wireless links than normal TCP; we have achieved throughput speedups of up to 20 times over regular TCP and handoff latencies over 10 times shorter than other mobile routing protocols.This work was supported by ARPA Contract J-FBI-93-153. This paper was in part presented at the ACM Mobile Computing and Networking Conference (Mobicom '95), Berkeley, California, 14–15 November 1995.     

An efficient transport service for slow wireless telephone links

Modern digital cellular telephones and portable computers have created a new platform for distributed information processing. However, the characteristics of wireless telephone links are different from those of wireline links. With standard TCP/IP protocols, this can lead to severe performance problems; some are related to the control of the wireless link, some to the cooperation of the wireless link and the fixed network. One possible solution is to split the end-to-end communication path into two parts, and to establish a separate control for each part. The Mowgli communication architecture is a sophisticated elaboration of this basic idea covering several data communication layers. One of its main components is the Mowgli data channel service (MDCS), which transparently replaces the standard TCP/IP core protocols over the slow wireless link. We discuss how the Mowgli approach, using the MDCS, alleviates the problems encountered with TCP/IP protocols over slow wireless links. The results of our performance tests indicate the merits of the Mowgli approach. The transfer times and the response times become more stable, transfer times for multiple parallel bidirectional transfers are substantially reduced, and response times in interactive work can be kept at a low and predictable level, even when there is other traffic on the wireless link     

Performance analysis of UDP with energy efficient link layer on Markov fading channels

In this paper, we analyze the throughput and energy efficiency performance of user datagram protocol (UDP) using linear, binary exponential, and geometric backoff algorithms at the link layer (LL) on point-to-point wireless fading links. Using a first-order Markov chain representation of the packet success/failure process on fading channels, we derive analytical expressions for throughput and energy efficiency of UDP/LL with and without LL backoff. The analytical results are verified through simulations. We also evaluate the mean delay and delay variation of voice packets and energy efficiency performance over a wireless link that uses UDP for transport of voice packets and the proposed backoff algorithms at the LL. We show that the proposed LL backoff algorithms achieve energy efficiency improvement of the order of 2-3 dB compared to LL with no backoff, without compromising much on the throughput and delay performance at the UDP layer. Such energy savings through protocol means will improve the battery life in wireless mobile terminals.     

Performance impact of interlayer dependence in infrastructure WLANs

Widespread deployment of infrastructure WLANs has made Wi-Fi an integral part of today's Internet access technology. Despite its crucial role in affecting end-to-end performance, past research has focused on MAC protocol enhancement, analysis, and simulation-based performance evaluation without sufficient consideration for modeling inaccuracies stemming from interlayer dependencies, including physical layer diversity, that significantly impact performance. We take a fresh look at IEEE 802.11 WLANs and using experiment, simulation, and analysis demonstrate its surprisingly agile performance traits. Our findings are two-fold. First, contention-based MAC throughput degrades gracefully under congested conditions, enabled by physical layer channel diversity that reduces the effective level of MAC contention. In contrast, fairness degrades and jitter increases significantly at a critical offered load. This duality obviates the need for link layer flow control for throughput improvement. Second, TCP-over-WLAN achieves high throughput commensurate with that of wireline TCP under saturated conditions, challenging the widely held perception that TCP throughput fares poorly over WLANs when subject to heavy contention. We show that TCP-over-WLAN prowess is facilitated by the self-regulating actions of DCF and TCP feedback control that jointly drive the shared channel at an effective load of two to three wireless stations, even when the number of active stations is large. We show that the mitigating influence of TCP extends to unfairness and adverse impact of dynamic rate shifting under multiple access contention. We use experimentation and simulation in a complementary fashion, pointing out performance characteristics where they agree and differ.     

Cross-layer speculative architecture for end systems and gateways in computer networks with lossy links

The throughput degradation of Transport Control Protocol (TCP)/Internet Protocol (IP) networks over lossy links due to the coexistence of congestion losses and link corruption losses is very similar to the degradation of processor performance (i.e., cycle per instruction) due to control hazards in computer design. First, two types of loss events in networks with lossy links are analogous to two possibilities of a branching result in computers (taken vs. not taken). Secondly, both problems result in performance degradations in their applications, i.e., penalties (in clock cycles) in a processor, and throughput degradation (in bits per second) in a TCP/IP network. This has motivated us to apply speculative techniques (i.e., speculating on the outcome of branch predictions), used to overcome control dependencies in a processor, for throughput improvements when lossy links are involved in TCP/IP connections. The objective of this paper is to propose a cross-layer network architecture to improve the network throughput over lossy links. The system consists of protocol-level speculation based algorithms at transport layer, and protocol enhancements at middleware and network layers that provide control and performance parameters to transport layer functions. Simulation results show that, compared with prior research, our proposed system is effective in improving network throughput over lossy links, capable of handling incorrect speculations, fair for other competing flows, backward compatible with legacy networks, and relatively easy to implement.     

Cross-Layer Enhancement to Support TCP-Based Traffics in WLANs

Widespread deployment of wireless local area networks and a gradual increase in streaming applications have brought about a demand for improved quality of service (QoS) in wireless networks. However, increasing user datagram protocol based high priority multimedia traffic and the class differentiation introduced in QoS protocols, has resulted into transmission control protocol (TCP) starvation and increased spurious timeouts. While today’s Internet traffic is still dominated by TCP based applications, the negative effects of IEEE 802.11e enhanced distributed coordination function (EDCF) scheme on TCP performance in the presence of high priority traffic have not been extensively explored. In this paper, the performance of TCP in 802.11e WLAN competing with high priority traffic is examined. The prioritised adaptive enhanced scheme (PAD_EDCF) is proposed. The proposed scheme gives priority to TCP control packets in order to improve the low traffic transmission flow and acquires additional capability of adjusting the MAC parameters based on the traffic load condition. Simulation results demonstrate that the proposed scheme significantly improves TCP performances in terms of traffic efficiency, throughput and reduces delay.     

Stabilizing TCP performance over bursty wireless links through the combined use of link-layer techniques

Improving TCP performance over wireless fidelity (WiFi) networks is recognized as a crucial issue, specially in those links prone to suffer from high bit error rate (BER) and bursty error losses. This paper shows how TCP throughput over a real IEEE 802.11b wireless local area network (WLAN) in a lossy office environment is stabilized by a combination of an adaptive forward error correction (FEC) scheme in conjunction with a Snoop agent, both added to the idle repeat request (RQ) mechanism inherently used by such technology. The most innovative aspects of this work are that these techniques operate simultaneously and they have been tested over a real scenario.     

ROBUST DATA LINK PROTOCOLS FOR CONNECTION-LESS SERVICE OVER SATELLITE LINKS

The performance of the connection-less network service (CLNS) over a satellite link with features such as high propagation delay and a varying (often high) level of errors, can be improved using the connection oriented data link (CODL) service. However, although existing CODL protocols satisfy the requirements of a connection oriented network protocol (e.g. X.25), they are not optimal for the CLNS over satellite links. The interaction of link error recovery procedures with the error recovery procedures implemented by CLNS users, the unnecessary delay due to link protocol sequencing and recovery procedures, and the interaction between applications with different quality of service (QoS) requirements cause most CODL protocols to have a detrimental impact on the performance of the CLNS. This study presents a new link protocol, which improves the quality of the CLNS (e.g. IP), especially considering the enhancement of the performance of connection-oriented transport protocols (e.g. TCP). The performance improvement using the new link protocols is demonstrated by implementation in a satellite interworking unit.