Recent simulation studies of TCP congestion control Slow-Start clearly show problems of oscillatory behavior in packet delay, congestion windows and throughput. Furthermore, it also suffers from rapid fluctuations in queue length, periodic packet losses, and underutilization of the transmission line in a large pipeline size and/or two-way traffic environments. These studies also exhibit three fairness problems in TCP-SS: systematic discrimination of a particular connection in periodic traffic, a bias against a connection with a long round trip time and a bias against bursty traffic. Based on the principles of rate control and preventive congestion mechanism, the authors have developed a new adaptive rate-based (ARB) congestion avoidance scheme that avoids these problems present in TCP-SS. It is demonstrated through simulations that ARB outperforms TCP-SS in network efficiency and fairness.>

Two local access algorithms are presented for a hierarchical mobile packet radio network where each user transmits with a fixed or an adjustable transmission radius, respectively. A model is developed for analyzing the two access schemes, and it is shown that the network can achieve better performance if each node uses an adjustable transmission radius. Previous research assumed a static two-dimensional Poisson model to describe the distribution of users and cannot fully characterize the effect of user mobility. The proposed model accounts for this effect. Numerical examples are presented.>

A model is proposed for analyzing the performance of a mobile packet radio network. In this network, a node is allowed to move around with a random velocity. Hence, the distribution of the number of nodes in one's transmission circle is dependent on the mobility of the nodes. A queuing model is developed to evaluate this distribution and the effects of node mobility on system performance. This effect cannot be obtained by classical two-dimensional Poisson model. It is found that an optimal power radius is dependent on the node mobility.>

This paper studies the behavior of a packet switch which provides finite waiting space and receives packetized messages. The arrivals of the messages constitute a Poisson process. Each message consists of a random number of packets. The number of packets contained in a message is assumed to be an integer-valued random variable which may follow any arbitrary probability distribution. All packets residing in the buffer receive service from a single output transmitter operating synchronously at a constant rate. Each packet receives the same fixed service time from the transmitter and then leaves the system. Upon the arrival of a message, if the remaining buffer space is not enough to accommodate all packets of the message, then the entire message is completely rejected. Results such as message blocking probability, packet blocking probability, throughput, and mean delay have been obtained. Two different approaches, minislot approximation and the application of the residue theorem, are used to obtain these results. Especially, this combinatorially very complex problem is successfully solved by the residue theorem in a recursive manner. These results are useful in evaluating the performance of a packet switch. They are also useful for design purposes.