Feiling Jia

Scheduling variable-length messages in a single-hop multichannel local lightwave network

The design of a medium access control scheme for a single-hop, wavelength-division-multiplexing-(WDM) multichannel local lightwave network poses two major difficulties: relatively large transmitter/receiver tuning overhead and large ratio of propagation delay to packet transmission time. Most schemes proposed so far have ignored the tuning overhead, and they can only schedule fixed-length packet transmissions. To overcome these two difficulties, the authors propose several scheduling algorithms which can reduce the negative impact of tuning overhead and schedule variable-length messages. A separate channel (control channel) is employed for transmission of control packets, and a distributed scheduling algorithm is invoked at each node every time it receives a control packet. By allowing the length of messages to be variable, a long message can be scheduled with a single control packet transmission, instead of fragmenting it into many fixed-length packets, thereby significantly reducing the overhead of control packet transmissions and improving the overall system performance. Three novel scheduling algorithms are proposed, varying in the amount of global information and processing time they need. Two approximate analytical models are formulated to study the effect of tuning time and the effect of having a limited number of data channels. Extensive simulations are conducted. Average message delays are compared for all of the algorithms. >

The receiver collision avoidance (RCA) protocol for a single-hop WDM lightwave network

We propose a multiple access protocol for a single-hop wavelength division multiplexing (WDM)-based multichannel local lightwave network. The protocol can support a large number of high-speed bursty traffic nodes interconnected via a passive optical star coupler. Each node is equipped with only one transmitter and one receiver, both of which are tunable over all the channels. A single control channel is established to arbitrate access to the other channels, called data channels. Access to the control channel is provided via a variation of slotted ALOHA so that the system is scalable. Unlike other protocols proposed in [11], [18], [23], our protocol contains a simple mechanism which can dynamically detect and avoid receiver collisions (hence it is called the receiver collision avoidance (RCA) protocol)

Variable-length message scheduling algorithms for a WDM-based local lightwave network

Two major difficulties in designing a single-hop multichannel local lightwave network are: relatively large transmitter/receiver tuning overhead and large ratio of propagation delay to packet transmission time. The authors propose several scheduling algorithms which can reduce the negative impact of tuning overhead and schedule variable-length messages. Thus, a long message can be scheduled with a single control packet transmission, instead of being segmented into many fixed-length packets, thereby significantly increasing the systems efficiency. Three novel scheduling algorithms are proposed, varying in the amount of global information and processing time. Two approximate analytical models are formulated to study the effect of tuning time and the effect of having a limited number of data channels. The systems performance is found to improve (1) if a simple mechanism is employed to avoid unnecessary transceiver tuning and/or (2) if a predictive transmitter tuning strategy is adopted.<<ETX>>

A high-capacity, packet-switched, single-hop local lightwave network

A new network architecture, called MultiS-Net, is proposed to provide packet-switched services to a large number of bursty users connected to a multichannel broadcast medium in a local environment. An example physical network would consist of a passive star coupler to which nodes are connected via two-way fibers, and wavelength division multiplexing (WDM) is used to create multiple, parallel optical channels. A separate channel, called a control channel, is established to coordinate data packet transmissions on the other channels (called data channels). To allow low network interfacing costs, each node in MultiS-Net is only equipped with a single tunable transmitter and a single tunable receiver, both of which are responsible for data channel access as well as control channel access, and both of which are rapidly tunable. MultiS-Net incorporates a simple medium access mechanism, and it can also achieve high system throughput. In addition, MultiS-Net is scalable, i.e. it can accommodate a variable number of nodes as well as time-varying load.<<ETX>>

Bimodal throughput, nonmonotonic delay, optimal bandwidth dimensioning, and analysis of receiver collisions in a single-hop WDM local lightwave network

A wavelength division multiplexing (WDM)-based local lightwave network employing a passive-star topology is considered. Users access the control channel according to the slotted-ALOHA random access protocol and the data channels according to a delayed-ALOHA protocol. It is shown that this system can have a bimodal throughput characteristic if its control channel is over-dimensional (which means that the control channel is allocated a larger share of the total system bandwidth than is needed). A recommendation for proper network-bandwidth dimensioning is made. A finite user population model that enables one to analyze the degradation of the system throughput because of receiver collisions, which could be ignored in infinite population models, is developed. The systems delay characteristic is also analyzed, and, for the case of an overdimensioned control channel and small values of backoff delay (which is required to resolve packet collisions in a random access system), the mean packet delay is found to be nonmonotic, i.e., the delay can decrease even when the offered load is increased.<<ETX>>

Performance analysis of a generalized receiver collision avoidance (RCA) protocol for single-hop WDM local lightwave networks

A single-hop wavelength division multiplexing (WDM) based local lightwave network employing a passive star topology is considered. The system consists of a single control channel and a number of data channels. Each station is equipped with a transmitter and a receiver, both of which are tunable over all the channels. Nodes employ the control channel to arbitrate (coordinate) their access to the data channels. The attractiveness of this architecture is its extreme simplicity. We have previously proposed a Receiver Collision Avoidance (RCA) protocol for such a system in which all nodes are equidistant from the passive star. Under the RCA protocol, access to the control channel is provided via a variation of slotted ALOHA, which includes a simple mechanism that can dynamically detect and avoid receiver collisions. The protocol is scalable and can support a large number of bursty nodes with a relatively small number of data channels. In this paper, we consider an extended RCA protocol (E-RCA) to incorporate nonuniform distances, while maintaining all of the protocols original attractive features. The analytical model for the E-RCA protocol is difficult to formulate; therefore, extensive simulations were conducted under various distance distributions. Results indicate that the E-RCA protocol performs almost as well as the RCA protocol under the same average distance conditions. Also, like the RCA protocol, the E-RCA protocol is simple, based on practical assumptions, and can be readily implemented with current lightwave technology.

The superchannel scheme for integrated services on multiple access broadcast networks

Abstract Integration of packet voice with data on a multiple access broadcast network (MABN) is an important problem. Several protocols, e.g., real-time control, movable-boundary and Global Queue schemes have been proposed to integrate voice and data traffic on MABNs. Both the movable-boundary and the Global Queue schemes require modifications to existing data stations, and they do not allow voice stations to utilize unused data bandwidth. Also, under the real-time control protocol, data performance will degrade significantly if voice traffic is heavy. In this paper, we propose a scheme for integrated services on MABNs based on the “superchannel” concept and describe a protocol for voice/data integration. The superchannel is a logical channel which is established on existing MABNs, and each such channel can carry a different type of traffic under a special control protocol. The proposed scheme is fully compatible with existing MABNs, i.e., data stations can operate using their original protocol. By introducing a priority boundary, the channel utilization is improved under heavy voice traffic and light data load, while still maintaining a desired level of performance if the combined voice/data traffic load is heavy. Since interference of voice and data traffic only appears at the boundary of the superchannel cycle, the protocol can be relatively easily extended for multiple traffic types. We first describe various operations that need to be performed by voice stations under the superchannel scheme. Then, approximate mathematical models are formulated to analyze the channel overhead, voice loss rate, and data queue length. Finally, various analytical and simulation results are obtained to illustrate some features of the proposed scheme, and these results are compared with previous work.

Optical interconnects for multiprocessor architectures using wavelength-division multiplexing

Multiprocessor architectures based on optical interconnects employing wavelength division multiplexing (WDM) are considered. WDM is a technique used to divide the tremendous bandwidth of a single strand of fiber into many non-interfering wavelengths. System components (processing, memory, and/or I/O elements) can use these wavelengths as communication channels. The authors present the background needed to understand WDM architecture, outline several categories of single-hop communication protocols, and show how a multihop WDM-based multiprocessor can reconfigure itself to any virtual topology.<<ETX>>

MultiS-Net: a high-capacity, packet-switched, multichannel, single-hop architecture and protocol for a local lightwave network

A variety of multichannel single-hop network architectures and protocols have recently been proposed to exploit the vast capacity of fiber-optics which can well exceed the maximum electronic speed limit. Our MultiS-Net proposal is a new multichannel, single-hop architecture and protocol for a local network which provides packet-switched services to a large number of users connected to a multichannel broadcast medium. A separate channel, called control channel, is established to coordinate data packet transmissions on the other channels (called data channels). To allow low network interfacing costs, each node.in MultiS-Net is only equipped with a single tunable transmitter and a single tunable receiver, both of which are responsible for data channel access as well as control channel access. Without a separate transmitter and a separate receiver dedicated to the control channel, a node cannot have global knowledge of the network status all the time. Hence, global reservation is infeasible here, and all previously proposed schemes with the same low-cost network interface as in MultiS-Net fail to achieve a system throughput as high as those achieved by global-reservation schemes. MultiS-Net successfully incorporates a simple medium access mechanism, which can also achieve high system throughput. In addition, MultiS-Net can accommodate a variable number of nodes as well as time-varying load; typically, however, the number of nodes may be much larger than the number of available data channels. We analyze the performance of MultiS-Net by first constructing a multidimensional Markov chain. Then, the equilibrium point analysis (EPA) technique is adopted to obtain the average data packet delay and system throughput, which are verified by simulation. Numerical results from both the analytical model and simulations are presented, and the impact of various parameters on the system performance is discussed. Nearly-full utilization of total data channel bandwidth can be achieved under proper configurations of system parameters.