Journal of Interconnection Networks

There are many discussions on agreement protocols of multiple peer processes (peers) where every peer just aims at agreeing on one value out of values shown by the peers. In meetings of human societies, agreement procedures are so flexible that persons can change their opinions and can use not only all-condition where every person agrees on one value but also various types of agreement conditions like majority-condition. In this paper, we discuss a flexible agreement protocol of multiple peers by taking into account human behaviors in social agreement procedures. Each peer first takes a value v₁ and notifies the other peers of the value v₁. A peer p_i in turn receives values from other peers. Unless a set of the values from all the peers satisfy the agreement condition, the peer p_i can take another value v₂. In order to model the social human behavior, in our previous work we discussed E- and P-precedent relations and on values of a peer p_i. The relations and show that a peer p_i can take a value v₂ after taking a value v₁ and prefers v₁ to v₂, respectively. If a peer autonomously takes values based on its precedent relations, the peers might not make an agreement even if there exists a satisfiable set of previous values. We discuss what previous values the peer can take again. In this paper, we try to find a satisfiable set of previous values in a history of values which the peers have so far taken, in addition for each peer to taking a new value at each round.

We propose to use Inter Vehicle Communications as part of a distributed system for ubiquitous health monitoring of patients. The biomedical data will be collected by wearable health diagnostic devices, which include various types of sensors. When patients are traveling on highways, we propose to transmit the measured biomedical data towards the corresponding Health Monitoring Centers via an Inter Vehicle Communication Network, which will guarantee the needed reliability, especially in areas where other wireless communications are not available. In particular, it is important that the Inter Vehicle Communication Network provides different priorities to various types of messages, depending on their urgency. We extend, a previously proposed inter vehicle hierarchical routing protocol, with priority scheduling to provide the needed priority services. We study various scheduling techniques and compare their effects on the service.

In mobile ad hoc networks (MANETs), data replication is essential for improving data availability. In our previous work, we proposed a few updated data dissemination methods to refresh old replicas efficiently in MANETs. These methods disseminate updated data items every time when owners of original data items update the items or every time two mobile hosts are newly connected with each other and this causes heavy traffic in the entire network. In this paper, we assume applications that periodically execute read operations with strict deadlines to data items and propose several alternative updated data dissemination methods. These methods reduce the traffic for data dissemination while keeping a high success ratio for read operations.

Single-constrained QoS routing protocols have inherent defects when applied into wireless ad hoc networks. Due to a single constraint parameter is only considered, they can't always cope with the problems caused by the uncertainty of ad hoc networks well. They are not robust enough. In order to overcome the drawbacks of single-constrained QoS routing protocols and improve the Quality of Service (QoS) of ad hoc networks, this paper proposed a multi-constrained QoS routing protocol based on fuzzy logic. It is developed from Dynamic Source Routing (DSR). The proposed protocol is service-aware in the sense that it considers the QoS required by different types of services and takes different network state parameters as the constraint conditions for fuzzy based routing system. New route discovery procedure and novel route maintenance mechanism are designed to support corresponding QoS requirements. Speed of packets sending is also adjusted adaptively referring to the outputs of the proposed fuzzy system. Performance of the fuzzy based DSR protocol is measured and evaluated under different conditions. Simulation results show that the improved protocol has better QoS guarantee capabilities compared to single-constrained QoS routing protocols for large-scale networks in terms of lower delay, smoother delay variation and lower packet loss rate.

Energy efficiency is a key issue in wireless ad hoc and sensor networks. In order to maximize network lifetime, several directions have been explored, among them energy efficient routing. In this paper, we show how to extend the standardized OLSR routing protocol, in order to make it energy efficient. This extension selects the path minimizing the energy consumed in the end-to-end transmission of a flow packet and avoids nodes with low residual energy. As it has been shown that two-path routing is energy efficient, we compare this extension with a two-path source routing strategy (with different links or different nodes). Moreover, to take into account residual node energy, the native selection of multipoint relays of OLSR is changed. Three selection algorithms based on the minimum residual energy are evaluated. An extensive performance evaluation allows us to choose for EOLSR (Energy efficient OLSR) the best variant maximizing both network lifetime and delivery rate.

Network on-Chips (NoC) have emerged as an attractive communication architecture to address the increasing demand for communication in performance-sensitive applications. Employing differentiated services is a widely used approach to support QoS (Quality of Service) by relatively prioritizing traffic in the networks employing connection-less communication mechanisms. In this paper we present an analytical evaluation of the average message latency for wormhole-routed interconnect architectures exploiting a traffic prioritization mechanism to achieve differentiated services-based QoS. To verify the validity of the model we apply the method to the Spidergon NoC and compare the model against the results obtained from a discrete-event simulator developed using OMNET++.

Achieving efficient and fairness bandwidth allocations among connections in high bandwidth delay product networks has been a daunting challenge. Addressing the deficiency of previous protocols, this paper presents a novel spectrum of Epoch Convergent Congestion Control protocols for deployment in high speed and long distance networks. Most of the end to end high speed protocols that has been presented can be seen as a member of ECCC. Through extensive analysis of ECCC, we focus on SQUARE, one of the protocols in Epoch Convergent Congestion Control. Without the need to revise the current end-to-end architecture, SQUARE is shown to be efficient when there is bandwidth available, to be fair when many flows compete with each other for the same bottleneck, to be friendly when deployed with the conventional TCP, to be robust when there are oscillations in the network.

A wireless mobile sensor network is a collection of wireless mobile hosts forming nodes that are communicating without the aid of any centralized administration or standard support services. Nodes are classified as sensor nodes and router. Some nodes act both as sensors and routers. While traditional MAC protocols must balance between throughput, delay, and fairness concerns, WSN MAC protocols place an emphasis on energy efficiency as well. Schedule-based MAC protocols have been proposed for WSN. A common theme through all these protocols is putting radios to a low-power "sleep mode" either periodically or whenever possible when a node is neither receiving nor transmitting. In this paper, we propose a novel cluster-based infrastructure creation protocol, namely: the Neighbor-Aware Clusterhead (NAC). We investigate the performance of the NAC protocol under different sleep scheduling protocols and compare it against another cluster-based protocol. In NAC protocol nodes are synchronized with their clusterheads and are allowed to go sleep mode in order to conserve their energy without degrading the performance of the network. The network is divided into clusters managed by a clusterhead. The simulation results show that NAC protocol has a good potential to be a good candidate protocol for SNET.

Most query search on road networks is either to find objects within a certain range (range search) or to find K nearest neighbors (KNN) on the actual road network map. In this paper, we propose a novel query, that is, incremental k nearest neighbor (iKNN). iKNN can be defined as given a set of candidate interest objects, a query point and the number of objects k, find a path which starts at the query point, goes through k interest objects and the distance of this path is the shortest among all possible paths. This is a new type of query, which can be used when we want to visit k interest objects one by one from the query point. This approach is based on expanding the network from the query point, keeping the results in a query set and updating the query set when reaching network intersection or interest objects. The basic theory of this approach is Dijkstra's algorithm and the Incremental Network Expansion (INE) algorithm. Our experiment verified the applicability of the proposed approach to solve the queries, which involve finding incremental k nearest neighbor.

Traffic grooming in a WDM network consists of assigning to each request (lightpath) a wavelength with the constraint that a given wavelength can carry at most C requests or equivalently a request uses 1/C of the bandwidth. C is known as the grooming ratio. A request (lightpath) needs two SONET add-drop multiplexers (ADMs) at each end node; using grooming, different requests can share the same ADM. The so called traffic grooming problem consists of minimizing the total number of ADMs to be used (in order to reduce the overall cost of the network). Here we consider the traffic grooming problem in WDM unidirectional rings which has been recently shown to be APX-hard and for which no constant approximations are known. We furthermore suppose an all to all uniform unitary traffic. This problem has been optimally solved for specific values of the grooming ratio, namely C = 2, 3, 4, 5, 6. In this paper we present various simple constructions for the grooming problem providing approximation of the total number of ADMs with a small constant ratio. For that we use the fact that the problem corresponds to a partition of the edges of the complete graph into subgraphs, where each subgraph has at most C edges and where the total number of vertices has to be minimized.

Current studies of "messy" broadcasting have so far concentrated on finding worst-case times. However, such worst-case scenarios are extremely unlikely to occur in general. Hence, determining average-case times or tight upper bounds for completing "messy" broadcasting in various network topologies is both necessary and meaningful in practice. In this paper, we focus on seeking the average-case "messy" broadcast times of stars, paths, cycles, and d-ary trees, and finding good upper bounds for hypercubes. Finally, we derive a recursive formula to express the average-case time for a specific "messy" broadcast model on a complete graph using a classical occupancy problem in probability theory, and provide a nice simulation result which indicates that this model behaves like classical broadcasting.