Hiroshi Inai

A two-layer queueing model to predict performance of packet transfer in broadband networks

In this paper, we propose an approximate analytical method to predict the performance of packet transfer in broadband networks. Since the ATM is a promising way to achieve a multimedia data transmission in broadband networks, the underlying network is assumed to be an ATM network. To represent the packet-level and the cell-level data transfers, we introduce a two-layer queueing network model. In the higher layer of the model, the source node (the sender) generates packets and then segments the packets into cells, while the destination node (the receiver) reassembles the arriving cells into the packets. The lower layer model is a tandem queueing network in which each queue corresponds to an ATM switch. We employ a discrete-time analysis of the queueing network model to obtain the packet-level performance measures, such as the end-to-end packet transfer delay, the packet loss probability, and the throughput. In numerical examples, we show some packet transfer delay distributions and effects of the packet size on the packet loss probability.

Dynamic threshold control in an ATM switch with shared memory buffer

To achieve fair sharing of buffer memory among the output port queues in a shared buffer ATM switch, a maximum allowable queue length (a threshold) for each queue should be set. This paper proposes a new dynamic queue length threshold control scheme in which a simple function is used to calculate the threshold value. Via simulation, we investigate cell loss characteristics when a single class and two classes of cells arrive under nonuniform load conditions. The results show that the threshold should be set to a slightly larger value than the current vacant buffer spaces to reduce the cell loss probability.

A request cluster allocation method for DOCSIS CATV/HFC networks

As an access medium to the Internet, CATV/HFC networks have been widespread recently. In such networks, each station reserves its data PDU transmission during a request period called a request cluster. DOCKS, a de facto standard for the networks, employs the binary back-off method for request cluster allocation. Since that method normally allocates unnecessary large request cluster, the transmission delay increases under heavy load conditions. We propose a request cluster allocation method which dynamically changes the cluster size according to the load conditions. In order to evaluate performance of the proposed method, we build a queuing model and execute computer simulation. The simulation result shows that the proposed method provides smaller delay than the binary back-off method.

Performance evaluation of a request cluster allocation scheme for multi-channel DOCSIS

DOCSIS is a de facto standard for data transmissions on CATV/HFC networks. In DOCSIS, an upstream channel is shared by many stations, and the stations send out their data PDUs according to the random reservation manner. Before data transmission, the station must send a request in a random access manner during the reservation period called a request cluster. Recently, DOCSIS has introduced a concept in that the multiple upstream channels can be combined together to get greater aggregated bandwidth. However, DOCSIS does not provide a request cluster allocation scheme for the aggregated upstream channel. On the other hand, for single upstream channel, DOCSIS employs the binary exponential back-off scheme for request cluster allocation. But, that provides too small/large request periods which are used in the random reservation, the delay increases rapidly as the offered traffic increases. In this paper, we propose a request cluster allocation scheme for multichannel CATV/HFC networks. Our scheme dynamically changes the request cluster size according to the traffic conditions. In order to evaluate performance of the proposed scheme, we build a queueing model and execute computer simulation. Simulation results present that the proposed scheme provides the smaller delay than the conventional scheme.

Performance evaluation of CSK spread spectrum Boomerang system

In this paper, the spread-spectrum (SS) Boomerang system using code-shift-keying (CSK) is proposed. CSK is one of the orthogonal modulation systems which transmit an orthogonal code sequence according to information bits. The conventional SS-Boomerang system which used direct-sequence spread-spectrum (DS/SS) technique applies the spreading sequence to every information bits. However, since the proposed SS-Boomerang system uses CSK, it can transmit more than one information with one spreading sequence. It is able to improve information transmission rate. Moreover, it is shown that the bit error rate and distance ranging performance of the proposed system is superior to that of the conventional system.

A back-off scheme for IEEE802.11 wireless LANs

Conventional IEEE802.11 wireless LANs employ the binary exponential back-off scheme to determine a back-off window size. Performance evaluations of the binary exponential back-off scheme and several enhanced ones have been done under an assumption that each station can store away only one packet with its transmission buffer. Under such assumption, enhanced schemes provide better performance, however they do not work well under a realistic condition that each station has more larger transmission buffer. In this paper, we propose a back-off scheme in which each station changes the back-off window size according to the load conditions. In order to evaluate performance of the proposed scheme, we build a queueing model and execute computer simulation. Simulation results show that the proposed scheme provides a higher throughput than the conventional ones. Moreover, under the proposed scheme, the number of stations has a slight influence on the performance.

Prioritized Service for High-Speed Terminals in Wireless LANs

In a wireless LAN, due to a radio wave condition and/or coexistence of several specifications, there exist some terminals of various transmission rates. Under the standard IEEE 802.11b/g protocol, all the terminals can achieve the same throughput since they equally obtain the channel access opportunities. In such a case, high-speed terminals cannot show their transmission abilities. To solve that problem, this paper proposes a channel access method which provides the access privileges to the high-speed terminals. Under an employment of the proposed method, the access point rejects the RTS frames origined from the low-speed terminals in a certain condition. Via simulation, the proposed method achieves appropriate throughput to each terminal. Furthermore, as compared with the standard and previously proposed methods, our method provides smaller file transfer delay for low-speed terminals.

An approximate transient analysis of dynamic thresholds in an ATM switch with shared buffer memory

To achieve a fair share of buffer memory among the output queues in a shared buffer ATM switch, a maximum allowable queue length (a threshold) for each queue should be set. In this paper, to analyze a transient behavior of the buffer state in such switches, we perform a discrete-time transient analysis of a queuing model and then approximately derive time-varying queue length distribution and cell loss probability. In numerical examples, we discuss the validity of our analytical approach by comparison with simulation results.