Debessay Fesehaye Kassa

An analytical model of TCP performance

The transmission control protocol (TCP) is the dominant transport layer protocol for the end-to-end control of information transfer. Accurate models of TCP performance are a key and basic step for designing, dimensioning and planning IP (Internet Protocol) networks. Packet-level simulation models do not scale with the growth of network capacities and number of users. Measurements can only be used to explore existing network scenarios or otherwise they become costly and inflexible with the growth and complexity of the Internet. This study presents a simple, fast and accurate analytical model of TCP. The model gives Internet performance metrics, assuming that only basic network parameters such as the network topology, the number of users, link capacity, distance between network nodes and router buffer sizes are known. The TCP performance model derives performance metrics which express the network quality of service. To obtain the performance metrics, TCP and network sub-models are used. A closed network of ./G/infin queues is used to develop each TCP sub-model where each queue represents a state of a TCP connection. An M/M/1/K queue is used for each network sub-model which represents the output interface of an IP router with a buffer capacity of K packets. The two sub-models are iteratively solved. Initial comparisons of our results with ns2 simulation experiments and with results from a well known TCP model show that our model is faster, simpler and gives more accurate TCP performance metrics

Convergence of the fixed point algorithm of analytical models of reliable internet protocols (TCP)

Analytical models are important tools for the performance investigation of the Internet. The literature shows that the fixed point algorithm (FPA) is one of the most useful ways of solving analytical models of Internet performance. Apart from what is observed in experimental literature, no comprehensive proof of the convergence and uniqueness of the FPA is given. In this paper we show how analytical models of reliable Internet protocols (TCP) converge to a unique fixed point. Unlike previous work in the literature the basic principles of our proof apply to both single and multiple bottleneck networks, to short and long-lived TCP connections and to both Drop Tail and Active Queue Management (AQM) routers. Our proof of convergence is based on a well known fixed point theorem and our uniqueness proof exploits the feedback nature of the reliable protocol. The paper specifies conditions under which the FPA of analytical models of TCP converges to a unique point. The concepts used in the proof can also be extended to analyze the equilibrium, stability and global uniqueness issues of TCP, other reliable protocols and the Internet as a whole.

An efficient queue management (EQM) technique for networks

In this paper I present a simple and efficient queue management (EQM) technique based on optimal bandwidth allocation for networks. Each router calculates the bandwidth share (throughput) of each interfering flow at a link using link capacity and Interface IP Address (see RFC 3630) or round trip time (RTT) and congestion window size of the flows which are carried in the packets. The sources send packets and the routers in the path drop or allow the packets based on a certain probability and label the non-dropped packets with the allowable optimal throughput for the next round. Each router in the path modifies this label and the sources eventually adjust their sending rates based on the allowable throughput they get from the returning ACKs. In this way EQM finds the fair bandwidth allocation and gives fair queue management. I also a prove that EQM can converge to a stable point.