We observe that even when the system throughput is relatively stable,end users of media-streaming applications do not necessarily experience smooth throughput, due to the unsynchronized window adjustments triggered by random congestion indications. We analyze and evaluate the negative impact of random window adjustments on smoothness, short-term fairness, and long-term fairness. We further propose an experimental congestion avoidance mechanism, namely TCP(/spl alpha/, /spl beta/, /spl gamma/, /spl delta/), based on coordinated window adjustments. The flow-level smoothness is enhanced significantly for media-streaming applications, without a cost on fairness and responsiveness. Responsiveness is even boosted when bandwidth is underutilized.

Additive increase/multiplicative decrease-based protocols, including transmission control protocol (TCP), TCP-friendly, and a new generation of rate-based protocols, attempt to control the tradeoff of responsiveness and smoothness. Traditionally, smoothness has not been a main concern since it does not impact the performance of regular Internet applications such as the Web, FTP, or e-mail. However, multimedia-driven protocols attempt to favor smoothness at the cost of responsiveness. In general, smoothness and responsiveness constitute a tradeoff; however, we uncover undesirable dynamics of the protocols in the context of wireless/mobile networks with high-error rate or frequent handoffs: low responsiveness is not counterbalanced by gains in smoothness, but instead, produces a conservative behavior that degrades protocol performance with both delay-tolerant and -sensitive applications. Based on our observations, as well as on further analysis of the impact of the bottleneck queue on channel utilization, we seek an alternative strategy for smooth window adjustments. We introduce a new parameter /spl gamma/, which implements a congestion avoidance tactic and reaches better smoothness without damaging responsiveness.

In this paper, we observe that although multiplicative decrease is necessary to accomplish fairness in congestion control, it does not inevitably sacrifice system throughput, as long as the system operates between the knee and the cliff, according to an equation. However, even when the system throughput is relatively stable, end users of real-time applications do not necessarily experience a smooth traffic, mainly caused by the unsynchronized window adjustments due to random congestion indications. We analyzed and evaluated the negative impact of random window adjustments on smoothness, short-term fairness, and even long-term fairness measured by a novel fairness metric defined in this paper. Therefore, we propose an experimental congestion avoidance mechanism to improve TCP smoothness for media-streaming applications. The mechanism relies on a fine-grained RTT estimation to measure the network condition, and coordinates the upward and backward window adjustments to abolish the damage of unsynchronized window control. Congestive packet drops are reduced by a new control parameter g, and the bottleneck queue length can also be controlled in an end-to-end way. Simulation results confirm that the new mechanism enhance significantly the smoothness and fairness, without a cost of responsiveness. In fact, by enabling a new parameter d, the responsiveness can be even enhanced when the bandwidth is under-utilized.

TCP(/spl alpha/,/spl beta/) protocols trade the congestion window increase value /spl alpha/ for the decrease ratio /spl beta/, to generate smoother traffic patterns and to maintain a friendly behavior. We study the design assumptions of TCP(/spl alpha/,/spl beta/) protocols and discuss the impact of equation-based modulation of /spl alpha/ and /spl beta/ on application efficiency. We confirm experimentally that, in general, smoothness and responsiveness constitute a tradeoff; however, we uncover undesirable dynamics of the protocols when the network or flow characteristics do not follow a prescribed and static behavior. For example, we show that smooth backward adjustments confine the protocol's capability to exploit resources that become available rapidly, and embarrass the fair and efficient growth of incoming flows. Furthermore, we show that in the context of wireless networks with high error rate, a low /spl alpha/ dictates a conservative behavior that degrades the protocol performance with both delay-tolerant and delay-sensitive applications; and in the context of high contention of heterogeneous flows, a low /spl alpha/ does not contribute to efficiency and friendliness.

We present a TCP-compatible and -friendly protocol which abolishes thr ee major shortfalls of TCP for reliable multimedia applications over heterogeneous networks: (i) ineffective bandwidth utilization, (ii) unnecessary congestion-oriented responses to wireless link errors (e.g., fading channels) and operations (e.g. handoffs), and (iii) wasteful window adjustments over asymmetric, low-bandwidth reverse paths. We propose TCP-Real, a high-throughput transport protocol that minimizes transmission-rate gaps, thereby enabling better performance and reasonable playback timers. In TCP-Real, the receiver decides with better accuracy about the appropriate size of the congestion window. Slow Start and timeout adjustments are used whenever congestion avoidance fails; however, rate and timeout adjustments are cancelled whenever the receiving rate indicates sufficient availability of bandwidth. We detail the protocol design and we report significant improvement on the performance of the protocol with time-constrained traffic, wireless link errors and asymmetric paths.