The queue size at a bottleneck would impact the performance of TCP protocols, especially when running a single TCP flow in networks with a large bandwidth-delay product. However, queue size has been not well considered in experiments. This paper shows how bottleneck queue size influences TCP protocols performance. Bursityness of advanced TCPs is examined. Ways of estimating queue size are introduced. Sending a UDP packet train until a loss is detected is a method to measure queuing delay to estimate queue size. Watching a loss in a TCP session to measure round trip time and calculate the queue size is also discussed. Results from experiments with a network emulator and a real network are reported. The results indicated that a layer-2 switch at a congestion point would be a major factor of decreasing TCP performance in a fast long distant path.

IEEE 802.11 standard and the enhanced amendments have defined several transmission rates for mobile stations to transmit and receive data frames. With the characteristic of modulation schemes, a higher network throughput can be expected by using higher level modulation scheme, but frame error probability will also become higher. Doubtlessly, it is an open issue of selecting a proper modulation scheme for a pair of mobile stations in time-varying indoor environment. This paper proposes a safe multiple access-rates transmission (SMART) scheme to enhance the reliability of data transmission in IEEE 802.11 multi-rate infrastructure wireless networks. SMART scheme provides reliable transmission by reserving a retransmission period which immediately follows the transmitted frame and is estimated by a lower transmission rate according to the transmitted frame. The performance of SMART scheme is evaluated by analytical model and simulation experiment, which show that the proposed scheme is significantly better than the IEEE 802.11 standard under the real environment with asymmetric traffic load. An enhanced SMART (ESMART) scheme is also proposed especially to improve bandwidth waste for long packet length at access point (AP). The performance measurements, such as goodput, throughput, average access delay, and service rate, are obtained and compared for the different schemes.

In recent years, requests for new advanced features, decentralization, and improvements in the forwarding speed, have been made for IX. A new IX architecture using MPLS (Multiple Label Switching) technology (MPLS-IX) is proposed to solve these requirements and rectify the problems. In this paper, a comparison between current IX and MPLS-IX based on router's packet forwarding performance is examined, and the feasibility of MPLS-IX is discussed. Moreover, the QoS requirements to provide IX service are discussed.

To understand the characteristics of a multimedia service, such as the large volume of data transfer and real-time constraints, it is necessary to have a performance evaluation tool for an HDD. Our HDD simulator is running on a PC operated on FreeBSD UNIX OS. We first investigate the seek time and the sustained rate of HDDs and then evaluate the performance of an HDD for an experimental VOD system. Applying the experimental results, we find the bottleneck of an HDD, and then suggest what HDDs are to be selected for a VOD system.

This paper studies a set of lip synchronization mechanisms for heterogeneous network environments. The set consists of four schemes, types 0 through 3, which are classified into the single-stream approach and the multi-stream approach. Types 0 and 1 belong to the single-stream approach, which interleaves voice and video to form a single transport stream for transmission. On the other hand, types 2 and 3, both of which are the multi-stream approach, set up separate transport streams for the individual media streams. Types 0 and 2 do not exert synchronization control at the destination, while types 1 and 3 do. We first discuss the features of each type in terms of networks intended for use, required synchronization quality of each medium, physical locations of media sources and implementation complexity. Then, a synchronization algorithm, which is referred to as the virtual-time rendering (VTR) algorithm, is specified for stored media; MPEG video and voice are considered in this paper. We implemented the four types on an ATM LAN and on an interconnected ATM-wireless LAN under the TCP protocol. The mean square error of synchronization, total pause time, throughput and total output time were measured in each of the two networks. We compare the measured performance among the four types to find out which one is the most suitable for a given condition of the underlying communication network and traffic.