Search

The quasi-synchronous transfer mode (Q-STM) method realizes cell-loss free and intranetwork delay jitter-free transmission on asynchronous transfer mode network, by allocating transmitted cell positions. In this paper, some design issues of the Q-STM method are considered and the applicability of the Q-STM method is investigated. First, the algorithms of the call setup procedure for Q-STM are compared. The effect of the number of subframes (NS) is investigated, and the most efficient NS in an assumed environment is obtained. Moreover, the QOS of ATM service class and processing time of a call setup are evaluated. Through these evaluations, we show guidelines for implementing the Q-STM and the applicability of the Q-STM.

In the asynchronous transfer mode (ATM) switch, unbalance with respect to the statistical destination port selection among the output ports can exist. There has been some research on output buffer-type ATM switching architecture in the case of the point-to-point cell arrival, assuming such nonuniform load condition of the output ports. However, the multicast traffic arrival condition, very important in future multimedia networks, has not been considered. Therefore, we evaluate the switch performance of the output buffer-type architecture in the case of including multicast cells at various ratios among the arrival cells on condition that the total amount of traffic load through all output ports is constant. As a result, it is clarified that exchanging the multicast cells intentionally in the ATM switch leads to the improvement of cell-loss performance in the output buffers having statistically high selection rate, if the buffers that are offered statistically high load, actually accommodate as many of the multicast cells as possible. From these results, by means of the network control that incorporates the multicast calls at a constant ratio in each ATM switch, it is possible to prevent the excessive load concentration to specific output port having statistically high utilization.

This paper considers a new transmission method (quasisynchronous transfer mode, Q-STM), as a means to eliminate both the cell loss and the network delay fluctuation in an ATM network. the concept of time-switch in circuit switching is introduced into the ATM switching framework. A switching structure and the call establishment procedure are proposed. In the proposed method, the time-switch is introduced, which eliminates in principle the output port contention in the ATM switching fabric. By preparing a buffer with a certain capacity, cell loss due to buffer overflow in the output port of the ATM switch can be completely eliminated.

This paper considers the shared output buffertype switching system as an architecture suited to the large-scale ATM switching system. an architecture is proposed, which can solve the fair throughput assignment problem among the ports belonging to the same shared memory, which is a problem in sharing the output buffer. the proposed architecture realizes fair access to the destination port contained in the specified shared buffer, by sort processing and the address attachment, without using the random access contention unit to the buffer in the knockout switch. For the proposed architecture and the conventional shared output buffer-type knockout switching system, the performances are compared by deriving the cell loss probability due to the output contention for each port. As a result, it is verified that the proposed architecture is better in terms of the realized cell loss probability. The performance of the switching system is evaluated by a computer stimulation when various buffering schemes are applied to the shared output buffer. Then it is shown that the fair throughput assignment among the considered output ports can be realized by combining the proposed architecture with an adequate buffering scheme.

The problem of accommodating multicast traffic in a virtual path (VP) environment under asynchronous transfer mode (ATM) conditions is examined. A network structure scheme suitable for all possible multicast communications is proposed. the expected result, statistical multiplexing effects during the accommodation of multiple multicast traffic over a VP should reduce hardware costs associated with the accommodation of multicast traffic. Multicast routing algorithms (the carrier-to-noise ratio (CNR) algorithm and the improved CNR algorithm) for an ATM network environment are proposed here. the efficiency of these algorithms is tested using their basic characteristics in computer simulations.

This paper proposes a prioritized resolvable-contention-ring scheme for high-speed optical communication networks and investigates its performance. The priority mechanism is based on the time-token control scheme. This mechanism attempts to guarantee the frame delay of time-critical traffic by controlling the frame transmission of noncritical traffic. The noncritical frame can be transmitted only if token rotation time is faster than target rotation time. This paper examines the priority mechanism in the case of two priority levels and also in the case of four priority levels. System performance is studied using a computer simulation method with a time-continuous simulator. The validity of the priority mechanism is demonstrated in a comparison with a fiber distributed data interface (FDDI) and the nonprioritized resolvable-contention ring. Efficient operation of the priority mechanism is investigated by considering the buffer overflow rate of each priority class. The proposed scheme performs effectively. Moreover, the performance of FDDI is comparatively degraded because of long token waiting time and large token capture time.

This paper discusses the channel assignment problem in the multichannel, high-speed packet network. It is assumed that multiple channels exist between a pair of switches in the network. It is expected that by regarding those channels as a single channel group and executing the packet transmission, a more efficient high-speed and large-capacity transmission will be realized. This paper newly introduces, in contrast to the virtual circuit in the multichannel network, the concept of a channel group-virtual circuit (CG-VC) for the multichannel network, where the packets are assigned to the channels based on the concept of the overflow traffic. To realize the CG-VC system in the high-speed environment, a simple and practical switch structure is proposed. The switch uses optical fibering as the basic module and the system is constructed by superposing the basic modules. Each packet automatically selects an appropriate channel from the set of output channels belonging to the same channel group. It is shown as a result of performance evaluation that, compared to the conventional VC packet switching system, the proposed system greatly improves the transmission delay and utilizes efficiently the channel.

This paper proposes two new access protocols, wherein the power control during transmission (PCX) scheme attempts improvement of the throughput by capture effect; this scheme is applied to s-ALOHA. Either method solves the problem of contradiction between the packet transmission time difference required in PCT scheme and the synchronized transmission as the basic requirement in PCT scheme and the synchronized transmission as the basic requirement in s-ALOHA. The first protocol is the s-ALOHA with capture interval (s-ALOHA/CI), which is concerned with the fixed-length packet format. A capture interval is provided in the first part of the slot, wherein the transmission time is selected at random to provide the transmission time difference. The second protocol is s-ALOHA with tail synchronization (s-ALOHA/TS), which is concerned with the variable-length packet format. Utilizing the variable-length property, the tail of the packet is synchronized to the tail of the slot and the packet is transmitted. Then the packet length difference produces the transmission time difference. It is verified through evaluation by analysis and simulation that the throughput performance of traditional s-ALOHA is greatly improved in either protocol by the capture effect. Especially, in contrast to the deterioration of the effective throughput of the traditional s-ALOHA due to the slot loss, it is verified that the performance of s-ALOHA/TS is better than that of the fixed-length packet scheme. These high-performance protocols will serve as a powerful means of realizing the effective operation of VSAT network, etc., with the synchronization function.

The advent of VSAT (very small aperture terminal) has made it possible to consider a new utilization of satellite communication. This paper proposes a new access protocol v-ALOHA, which is matched to the characteristics of the packet switching network based on VSAT. p-ALOHA is used as a means to cope with the requirement for simplification. To complement its disadvantage of low through-put, a new control scheme is combined based on the new utilization. The performance is improved by the capture effect. The basic idea of the control scheme is to vary the power with time while transmitting packets. The difference in the transmission-start time is converted into the power difference and produces the capture effect, thereby improving the performance. Thus, an access protocol is realized which improves the low throughput of p-ALOHA while preserving its simplicity. This paper describes the basic idea of this v-ALOHA and the protocol structure, indicating qualitatively its features. Then the throughput performance is formulated through a quantitative analysis. It is shown from the numerical results for the fixed-length packet that the performance of the proposed scheme is better than other capture schemes, approaching sufficiently the performance of s-ALOHA. For the case of the variable-length packet, there is no slot loss as in s-ALOHA, and the proposed scheme is shown, using an example, to be better sometimes than s-ALOHA.

This paper presents an analysis of the frame ALOHA scheme in the multiple-access broadcasting channel with the capture effect, which has not been taken into consideration in past analysis. First, the transmission-success rate, together with its expectation and variance, are determined, which indicate the basic performance of the group with various capture rates. The case is discussed where the capture rate is physically limited. Then two optimization problems are considered to maximize the number of terminals with access success. The first approach considers only a particular group, and a necessary condition is derived for the optimization of that group. Second, the performance of the whole system is improved, and the optimum solution is derived for the group assignment from the viewpoint of the capture rate. Finally, a discussion is made based on the numerical example, and it is shown that a higher throughput than in the past is realized in the multiple-access broadcasting transmission with capture effect. Furthermore, it is shown that a higher performance is realized if all terminals are optimized utilizing the capture effect.

A new configuration for packet radio networks, called local/global (LG) network configuration, is proposed, aiming at total throughput improvement. By using two kinds of channels (local channel and global channel) and geographically partitioning the total area, the spatial reuse of channel capacity is realized. The performance of this LG configuration is analyzed for a fixed radio network by introducing the delay ratio γ between local and global channels. s-ALOHA protocol is used under heavy overload. By the performance analysis and numerical results, it is concluded that the improvement of the total throughput is attained for a typical network configuration without greatly increasing the mean delay. Also, various schemes for determining the allocation ratio of channel capacity to the local and global channels are presented and studied to apply this LG configuration effectively.

A multichannel CSMA/CD scheme has been proposed which aims at utilizing the merits of CSMA/CD, a typical access control scheme in bus-type LAN, while overcoming the demerits. For this multichannel CSMA/CD scheme, the following three schemes have been proposed: node distribution, region distribution, and load distribution schemes. First, this paper compares and discusses the proposed schemes, especially the load and region distribution schemes. The effective range for each scheme is determined. Then a hybrid load distribution/region distribution scheme is proposed which aims at improving the performance and enlarging the range of applications by dynamically switching between the load distribution and the region distribution schemes, according to the load level of the network. Discussions and qualitative performance evaluations are presented for the protocol, realizability and reliability of the scheme. It is shown that the load distribution/region distribution scheme in the multichannel CSMA/CD scheme exhibits a very satisfactory performance over the entire range of the load, with excellent reliability and expandability.