End-to-end rate-based congestion control: convergence properties and scalability analysis

We study several properties of binary-feedback congestion control in rate-based applications. We first derive necessary conditions for generic binary-feedback congestion control to converge to fairness monotonically (which guarantees asymptotic stability of the fairness point) and show that AIMD is the only TCP-friendly binomial control with monotonic convergence to fairness. We then study the steady-state behavior of binomial controls with n competing flows on a single bottleneck. Our main result here shows that combined probing for new bandwidth by all flows results in significant overshoot of the available bandwidth and rapid (often super-linear as a function of n) increase in packet loss. We also show that AIMD has the best scalability and lowest packet-loss increase among all TCP-friendly binomial schemes. We conclude the paper by deriving the conditions necessary to achieve constant packet loss regardless of the number of competing flows, n, and, in both simulation and streaming experiments, examine one new scheme, called ideally scalable congestion control, with such constant packet loss.     

A study of deploying smooth‐ and responsive‐TCPs with different queue management schemes

While there exist extensive research works on congestion control and active queue management, or the joint dynamics of a congestion control strategy with the random early detection (RED) algorithm, little has been done on the interactions between different window adjustment strategies and different queue management schemes such as DropTail and RED. In this paper, we consider a spectrum of TCP‐friendly additive increase and multiplicative decrease (AIMD) parameters. At the one end of this spectrum, smooth‐TCP enhances smoothness for multimedia applications by reducing the window decrease ratio upon congestion, at the cost of the additive increase speed and the responsiveness to available bandwidth. At the other end, responsive‐TCP enhances the responsiveness by increasing the additive increase speed, at the cost of smoothness. We investigate the network dynamics with various combinations of AIMD parameters and queue management schemes, under different metrics. The investigation is conducted from the deployment (especially incremental deployment) point of view. We discussed the impact of the interactions on the goodput, fairness, end‐to‐end delay, and its implications to energy consumption on mobile hosts. Copyright © 2008 John Wiley & Sons, Ltd.     

Queue length‐based load balancing in data center networks

Traffic load balancing in data centers is an important requirement. Traffic dynamics and possibilities of changes in the topology (e.g., failures and asymmetries) make load balancing a challenging task. Existing end‐host–based schemes either employ the predominantly used ECN or combine it with RTT to get congestion information of paths. Both congestion signals, ECN and RTT, have limitations; ECN only tells whether the queue length is above or below a threshold value but does not inform about the extent of congestion; similarly, RTT in data center networks is on the scale of up to few hundreds of microseconds, and current data center operating systems lack fine‐grained microsecond‐level timers. Therefore, there is a need of a new congestion signal which should give accurate information of congestion along the path. Furthermore, in end‐host–based schemes, detecting asymmetries in the topology is challenging due to the inability to accurately measure RTT on the scale of microseconds. This paper presents QLLB, an end‐host–based, queue length–based load balancing scheme. QLLB employs a new queue length–based congestion signal that gives an exact measure of congestion along the paths. Furthermore, QLLB uses relative‐RTT to detect asymmetries in the topology. QLLB is implemented in ns‐3 and compared with ECMP, CONGA, and Hermes. The results show that QLLB significantly improves performance of short flows over the other schemes and performs within acceptable level, of CONGA and Hermes, for long flows. In addition, QLLB effectively detects asymmetric paths and performs better than Hermes under high loads.     

基于跳到跳信息的卫星网络传输控制协议研究

设计了一种适用于卫星网络的传输控制协议TPSN。该协议使用异步跳到跳确认,快速恢复成段丢失数据,并采用基于检测窗口的端到端选择性否定应答(SNACK)机制,减少协议控制信息,保证数据可靠传输。在跳到跳可靠性保证机制基础上传输网络负载,实现在避免网络拥塞条件下高效利用网络带宽资源,并保证具有不同端到端往返时延的异种数据流之间的公平性。仿真结果表明TPSN能够在长延迟、误码率高、链路频繁切换的卫星网络中保证高效的带宽利用、可靠的数据传输以及各数据流之间的公平。     

A low-state packet marking framework for approximate fair bandwidth allocation

Misbehaving, non-congestion-reactive traffic is on the rise in the Internet. One way to control misbehaving traffic is to enforce local fairness among flows. Locally fair policies, such as fair-queueing and other fair AQM schemes, are inadequate to simultaneously control misbehaving traffic and provide high network utilization. We thus need to enforce globally fair bandwidth allocations. However, such schemes have typically been stateful and complex to implement and deploy. In this letter, we present a low state, lightweight scheme based on stateless fair packet marking at network edges followed by RIO queueing at core nodes, to control misbehaving flows with more efficient utilization of network bandwidth. Additionally, with low-state feedback from bottleneck routers, we show that, in practice, we can approximate global max-min fairness within an island of routers. We show, using simulations, that we can indeed control misbehaving flows and provide more globally fair bandwidth allocation.     

基于RTT的TCP流带宽公平性保障机制

TCP端到端的拥塞控制机制使得TCP连接获得的瓶颈带宽反比于RTT(数据包往返时间)。为了缓解TCP对于RTT较小流的偏向,区分服务的流量调节机制在RTT较小的流取得目标速率且获得多余资源的情况下可以确保RTT较大流不至于饥饿。现有的基于RTT的流量调节机制在网络拥塞程度较轻时非常有效,但是当网络拥塞程度较重时,由于对RTT较大流的过分保护而导致RTT较小流饥饿。因此,通过引进自适应的思想提出了改进方法,其主要思想就是根据网络的拥塞程度自适应地调整对RTT较大流的保护程度。大量的仿真试验表明所提的机制能有效保障TCP流的带宽公平性并且比现有的方法具有更好的强壮性。     

Improving fairness of TCP Vegas

TCP Vegas exhibits fairness problems even for flows with the same round‐trip time (RTT). We propose an enhanced Vegas with three revisions, replacing BaseRTT with RTT, detecting how fast acknowledgements return and the acceleration of the return speed. The impacts of each of the three proposed revisions are not ignorable. The proposed novel Vegas with the three revisions, called EVA, achieves better fairness under various network conditions. Copyright © 2004 John Wiley & Sons, Ltd.     

A novel core‐stateless ABR‐like congestion avoidance scheme in IP networks

We propose a novel explicit rate‐based congestion avoidance scheme. The scheme is similar to ATM available bit rate (ABR) services but its complexity is largely reduced to enable implementation in IP networks. In our proposed scheme, sources can adapt their sending rate according to network status. Adaptation of sending rates converges to max–min fairness with minimal rate guarantee. Furthermore, routers do not maintain per flow state; they use FIFO packet scheduling enhanced by an explicit rate feedback mechanism with estimations, including number of active flows, number of bounded flows and total bounded rate, based on a simple algorithm motivated by Bloom filter. We present and discuss simulations on the performance under various network conditions. Copyright © 2005 John Wiley & Sons, Ltd.     

A Novel Congestion Detection Scheme in TCP Over OBS Networks

This paper introduces a novel congestion detection scheme for high-bandwidth TCP flows over optical burst switching (OBS) networks, called statistical additive increase multiplicative decrease (SAIMD). SAIMD maintains and analyzes a number of previous round-trip time (RTTs) at the TCP senders in order to identify the confidence with which a packet loss event is due to network congestion. The confidence is derived by positioning short-term RTT in the spectrum of long-term historical RTTs. The derived confidence corresponding to the packet loss is then taken in the developed policy for TCP congestion window adjustment. We will show through extensive simulation that the proposed scheme can effectively solve the false congestion detection problem and significantly outperform the conventional TCP counterparts without losing fairness. The advantages gained in our scheme are at the expense of introducing more overhead in the SAIMD TCP senders. Based on the proposed congestion control algorithm, a throughput model is formulated, and is further verified by simulation results.      

Contention‐based geographic forwarding in asynchronous duty‐cycled wireless sensor networks

In asynchronous duty‐cycled wireless sensor networks, it is desirable that the data forwarding scheme is adaptive to the dynamics caused by the uncertainty of sensor nodes’ working schedules. Contention‐based forwarding is designed to adapt to the dynamic environments. In this work, we are interested in the contention‐based geographic forwarding (CGF) for two asynchronous duty‐cycling (ADC) models, which we refer to as uninterruptible ADC (U‐ADC) and interruptible ADC (I‐ADC). We propose a new residual time‐aware routing metric for CGF in the I‐ADC model and present a residual time‐aware forwarding scheme using this metric. We evaluate the performance of CGF in both asynchronous duty‐cycling models. Simulation results show that CGF in the U‐ADC model provides a shorter delivery delay while suffering from a high sender effective duty cycle problem. CGF in the I‐ADC model incurs a very long data delivery delay, but it can achieve a good load balancing among nodes. It is also demonstrated that the proposed residual time‐aware forwarding scheme lowers the effects of the performance degradation caused by the pure asynchronous duty‐cycling operation. Copyright © 2011 John Wiley & Sons, Ltd.     

A fast‐convergent fairness scheme for resource allocation in RPR networks

In this paper, we propose a fast‐convergent fairness scheme in IEEE 802.17 resilient packet ring (RPR) networks. In the proposed scheme, each station could rapidly approach fair rate by estimating the number of unbounded flows at each link. In addition, the fast‐convergent scheme could prevent rate oscillations in the RPR aggressive mode scheme under unbalanced traffic. The estimation mechanism is simple and scalable since it is stateless without per‐flow management. Through analytical and simulation evaluations our scheme was found to be stable and speedy when compared with 802.17 RPR fairness scheme or the proposed distributed virtual‐time scheduling in rings scheme. Our scheme could allocate bandwidth fairly and smoothly among flows and achieve high utilization at the same time in the RPR network. Copyright © 2008 John Wiley & Sons, Ltd.     

A tradeoff resource allocation based on MF‐TDMA scheme in the multibeam data relay satellite systems

A tremendous increase in the number of distributed satellite constellations with the unscheduled burst data traffic will impose addition and diverse requirements on the DRS (data relay satellite) systems, which increases the complexity for beam management and affects a real‐time data return and acquisition. In this paper, we suggested that a large capacity can be achieved by a multibeam DRS system based on multifrequency time division multiple access scheme providing multiaccess for the distributed satellite constellations. Because the space‐based information network is characterized by the limited on‐board resources, a highly dynamic topology and time‐varying intersatellite links, we designed a 2‐stage dynamic optimization approach to separate the multiobjective optimization for frequency/time blocks and power, aiming at the rapidly converging to the optimal solution and at the same time meeting the fairness resource allocation. In particular, a capacity‐fairness tradeoff algorithm is proposed based on hybrid the enhanced genetic algorithm and the particle swarm optimization. Simulation results show that the tradeoff between maximizing total capacity and providing proportional fairness allocation is well balanced. The proposed algorithm can rapidly converge to adapt to the highly dynamic topology in data relay satellite systems.     

Fair QoS multi-resource allocation for uplink traffic in WLAN

In wireless local area network (WLAN), improving the quality of service (QoS) of users is often at odd with striking fairness among users. In this work, we suggest that in WLAN, multiple types of network resources should be jointly allocated to users to achieve “QoS fairness”, which is a new fairness concept targeting at balancing QoS and fairness in WLAN by allocating multiple types of network resources to users. To this end, we first transform user QoS requirements to multi-resource demands and apply the dominant resource fairness scheme to allocate network resources for each user. We prove several salient QoS-based fairness properties based on a model mapping between QoS and resources. We further discuss about more general conditions for diverse mapping models where QoS fairness properties can be satisfied. We find that the QoS fairness properties can be guaranteed as long as the mapping model meets a few practical requirements, indicating the wide applicability of our scheme. To consolidate our multi-resource allocation scheme, we design a practical protocol for WLAN. The simulation results validate that the QoS fairness can be guaranteed in practical WLAN scenario.     

End‐to‐end response time with fixed priority scheduling: trajectory approach versus holistic approach

In this paper, we are interested in providing deterministic end‐to‐end guarantees to real‐time flows in a distributed system. We focus on the end‐to‐end response time, quality of service (QoS) parameter of the utmost importance for such flows. We assume that each node uses a Fixed Priority scheduling. We determine a bound on the end‐to‐end response time of any real‐time flow with a worst case analysis using the trajectory approach. We establish new results that we compare with those provided by the classical holistic approach for flows visiting the same sequence of nodes. These results show that the trajectory approach is less pessimistic than the holistic one. Moreover, the bound provided by our worst‐case analysis is reached in various configurations, as shown in the examples presented. Copyright © 2004 John Wiley & Sons, Ltd.     

Fair Resource Allocation in Wireless Networks Using Queue-Length-Based Scheduling and Congestion Control

We consider the problem of allocating resources (time slots, frequency, power, etc.) at a base station to many competing flows, where each flow is intended for a different receiver. The channel conditions may be time-varying and different for different receivers. It is well-known that appropriately chosen queue-length based policies are throughput-optimal while other policies based on the estimation of channel statistics can be used to allocate resources fairly (such as proportional fairness) among competing users. In this paper, we show that a combination of queue-length-based scheduling at the base station and congestion control implemented either at the base station or at the end users can lead to fair resource allocation and queue-length stability.     

Core‐stateless fair rate estimation fair queuing

Core‐stateless mechanisms, such as core‐stateless fair queuing (CSFQ), reduce the complexity of fair queuing, which usually need to maintain states, manage buffers, and perform flow scheduling on a per‐flow basis. However, they require executing label rewriting and dropping decision on a per‐packet basis, thus preventing them from being widely deployed. In this paper, we propose a novel architecture based on CSFQ without per‐packet labelling. Similarly, we distinguish between edge routers and core routers. Edge routers maintain the per‐flow state by employing a fair queuing mechanism to allocate each flow a fair bandwidth share locally and a token bucket mechanism to regulate those flows with feedback packets sent from egress edge routers. Core routers do not maintain per‐flow state; they use FIFO packet scheduling extended by a fare rate alarm mechanism by estimating the arrival rate and the number of flows using a matching–mismatching algorithm. The novel scheme is called core‐stateless fair rate estimation fair queuing (CSFREFQ). CSFREFQ is proven to be capable of achieving max–min fairness. Furthermore, we present and discuss simulations and experiments on the performance under different traffic scenarios. Copyright © 2005 John Wiley & Sons, Ltd.     

Protocol design and data detection for two‐way relay networks with fractional asynchronous delays

In wireless two‐way relay systems, it is difficult to achieve perfect timing synchronization among different nodes. In this paper, we investigate relaying protocol design and data detect schemes for asynchronous two‐way relaying systems to combat the intersymbol interference caused by asynchronous transmission. We consider fractional asynchronous delays and two schemes are proposed based on cyclic prefixed single carrier block transmission, namely, the receiver frequency domain equalization scheme and relay synchronization and network coding (RSNC) scheme. In the receiver frequency domain equalization scheme, the relay simply amplifies the received signal and forwards to the two source nodes, and fractionally spaced frequency domain equalizer (FS‐FDE) is employed at the receiver to recover the transmit data. In the RSNC scheme, the asynchronous signals are resynchronized with an FS‐FDE at the relay node. The output signals of FS‐FDE are then demodulated and network coded before forwarding to the two source nodes. In this RSNC scheme, data detection at the source nodes is the same as that in synchronous networks because the asynchronous signals have already been synchronized at the relay node. Simulation results show that the performance of both schemes is almost the same as in the perfect synchronized two‐way relaying systems. Copyright © 2012 John Wiley & Sons, Ltd.     

A weighted fairness guarantee scheme based on node cooperation for multimedia WLAN mesh networks

In this paper, we focus on weighted fairness in multimedia WLAN mesh networks. Based on the analysis of the fairness problem of IEEE 802.11e Enhanced Distributed Channel Access (EDCA) scheme in WLAN mesh networks, we propose a weighted fairness guarantee scheme (WFGS), which provides weighted fairness for multimedia flows with different QoS requirements through node cooperation. WFGS piggybacks extra field in RTS/CTS frames to declare the channel occupation ratio of each flow. Accordingly, the transmitters can get the neighboring flows’ channel occupation ratio via overhearing the RTS/CTS frames from its neighbors, and cooperatively adjust the contention window size to achieve weighted fairness among the flows. Also, to reliably reserve transmissions, an adaptive power control based RTS/CTS handshake mechanism is introduced. Simulation results show that compared with EDCA scheme, WFGS can effectively resolve the collisions induced by the carrier interference and thereby guarantee both the short-term and long-term weighted fairness among multimedia flows.     

Fairness in MIMD Congestion Control Algorithms

The Mulitplicative Increase Multiplicative Decrease (MIMD) congestion control algorithm in the form of Scalable TCP has been proposed for high speed networks. We study fairness among sessions sharing a common bottleneck link, where one or more sessions use the MIMD algorithm. Losses, or congestion signals, occur when the capacity is reached but could also be initiated before that. Both synchronous as well as asynchronous losses are considered. In the asynchronous case, only one session suffers a loss at a loss instant. Two models are then considered to determine which source looses a packet: a rate dependent model in which the packet loss probability of a session is proportional to its rate at the congestion instant, and the independent loss rate model. We first study how two MIMD sessions share the capacity in the presence of general combinations of synchronous and asynchronous losses. We show that, in the presence of rate dependent losses, the capacity is fairly shared whereas rate independent losses provide high unfairness. We then study inter protocol fairness: how the capacity is shared in the presence of synchronous losses among sessions some of which use Additive Increase Multiplicative Decrease (AIMD) protocols whereas the others use MIMD protocols.