Search

Critical real time system must run under high dependable environment which guarantees a high level of reliability and security. Fault tolerance and security technique has played an important role in building such environment. In this paper, we provide an improved technique to schedule such tasks with high reliability and security requirement on high dependable distributed real time system. First, the paper analyzed the scheduling problem on high dependable distributed real time system and provided a refined scheduling model which accurately described processors, fault tolerant tasks and security overhead. Based on the scheduling model, it proposed a new and better scheduling algorithm--Security Driven Fault Tolerant Scheduling Algorithm (SDFT) which extended FTRMFF algorithm to tolerate N processors failure at one time, and it introduced a new global scheduler to improve efficiency of scheduling process. Moreover, the SDFT supported flexible security policy applied on real time tasks according to its security requirement and considered the effect of security overhead during scheduling. In addition, the paper adopted and extended the Completion Time Test (CTT) as the schedulability criteria. Simulation experiments showed the SDFT algorithm could achieve better reliability and security at the cost of less number of processors.

Data fusion or In-network processing methods were often adopted in Wireless Sensor Networks (WSNs) to reduce data communication and prolong network lifetime, which made WSNs application can be described as a set of tasks (sensing, processing) and dependencies among them. Task assignment has become an important problem which needed to be resolved for different tasks assignment caused different energy consumption. Based on the task graph of WSNs, an energy-efficient task assignment framework was proposed. As application task can be decomposed into sensing subtasks and processing subtasks, we present the task assignment as sensing subtask assignment and processing subtask assignment. We formulate the processing subtask assignment as a 0–1 quadratic programming problem and evaluate it through experiments.

Power control in wireless networks is important to maintain reliable communication links between base stations and nodes, and to maximize the battery life. An energy aware connectivity set based transmission power control protocol is presented in this paper. We augment our distributed protocol with a heuristic that considers energy cost information when making power control decisions. Compared to traditional protocol, our protocol can improve energy efficiency and prolong network lifetime significantly, while networks connectivity is still maintained in extreme scenarios. Simulation results demonstrate our conclusion.

In many applications of wireless sensor networks, sensor nodes are densely deployed in the area of interest. It is nontrivial to establish an efficient topology that provides communication under energy constrains and system dynamics. In this paper, a new hierarchical energy aware topology control protocol named HETCP is proposed. It combines energy balance metric and energy conserving metric together to optimize topologies. It forms sensor nodes into clusters to balance energy load and improve the network scalability. Inside clusters, a subset of nodes is selected to work while others remained sleeping in order to save energy. Both clustering and sleeping schedule algorithm are introduced in detail. Simulation results show that HETCP has excellent stability and energy efficiency

Wireless sensor networks lack some basic abilities taken for granted in wired networks, such as the ability to know adjacent nodes. Neighbor protocols are designed to exchange node information for determining which nodes are alive and reachable. This paper introduces the energy efficient adaptive neighbor detect protocol (EEAND), which automatically finds the balance point between energy cost and network connectivity. Towards traditional neighbor detect protocol, EEAND protocol only sends broadcast information when necessary to main topology. According to the SNR of the packets received from neighbor nodes, each node maintains its neighbor table. Simulation studies show that EEAND maintain similar delivery rates while using fewer control packets across a wide range of network characteristics.

A problem that must be faced when designing systems based on network processors (NPs) is how to assign the packet-processing tasks to the processor pipeline consisting of many processing engines. Nowadays, this work is done manually, which is error-prone and inefficient. To alleviate this, this paper proposes a genetic algorithm-based method to assign processing tasks to processing engines automatically. Experiments show that this method can quickly obtain near-optimal solutions from the large solution space and the algorithm execution time is decoupled with pipeline stages.

As network-centric computing becomes more pervasive and applications become more distributed, the demand for flexible and efficient message delivery is increasing. Messaging middleware is a key technique of message delivery between distributed computing applications. The main communication models of current messaging middleware are publish/subscribe and point-to-point. The two communication models have their own characteristics. However, neither of them can achieve efficiency and flexibility simultaneously. In this paper, we proposed and implemented a hybrid communication model based messaging middleware (HCM3), which takes full advantages of two models. Performance result shows that HCM3 can deliver message in heterogeneous environment with the features of high efficiency, flexibility and reliability.

The process of computational grid resource scheduling generally falls into two phases: global resource scheduling and local resource scheduling. Under this pattern, global resource scheduling algorithms are different from the traditional resource scheduling algorithms of LRMSs (local resource management system, LRMS). In this paper, a general self- adaptive global resource scheduling algorithm for Computational Grid, BMQOS (Best Multi QOS) is presented. According to the personal resource requirement of a computational grid job, BMQOS globally selfadaptively weighs every index of the MQOS of candidate computational grid nodes and chooses an appropriate node for a job from the candidates finally. Applied in NPU Campus Computational Grid, this algorithm achieves good effect.

This paper proposes the idea of integrating context aware computing theory with sensornet. Be context aware, sensor nodes are expected to use contexts available and adjust their behaviors correspondingly. This paper motivates and describes the idea of CASN (context aware sensornet). It also presents research challenges arising from CASN and outlines our research agenda.

To addressing the self adaptation problem arises from large scale densely deployed sensornet, we argue that integrating the principle of context aware with sensornet is feasible. To build such a context aware sensornet, proper context describing and storing mechanisms must be provided. In this paper, we propose RBLS, a Role Based Local Storage scheme. RBLS is designed simple and energy efficient. Aimed at providing context storage support for sensornet, RBLS stores contexts at node’s local space and dynamically allocates extra spaces according to the roles a node holding. A “snapshot” is used by RBLS to record a neighbor’s private contexts. We evaluate the performance of RBLS against a primitive scheme. Simulation results are included in this paper.

Advances in MEMS technology, wireless communications, and digital electronics have enabled the development of low-cost, low-power, multifunctional sensor nodes that integrating the ability of sensing, computing, and communication. The low per node cost will enable the development of densely distributed sensornet for a wide range of applications. Sensornet is usually deployed in remote and hostile area without human care. To achieve scalable, robust and long-lived goal, sensornet itself should execute adaptive adjustment and burn energy economically. In this paper we present the idea of context aware sensornet (CASN). Be context aware, sensor nodes are expected to use contexts available and adjust their behaviors correspondingly. Also, cooperative sensor nodes are expected to use context to promote their collaboration. The paper presents the motivation and research challenge of building CASN. It also briefly discusses the implementation issues of CASN middleware.

Distributed time service is a basic object service in distributed computing environment based on CORBA. We present a clock synchronization method based on clock groups. We divide machines in a system into groups, while nodes in a group are in the same LAN. There is a master clock and several slave clocks in a group, which carry out intra-group synchronization. Master clocks of all groups together accomplish inter-group synchronization. Such a dividing-to-group synchronization method makes the system scalable and flexible. At the same time, it reduces machines spanning LANs to attend synchronization, thus synchronization overhead is reduced. We have implemented clock synchronization based on clock groups in our experimental distributed object environment.

CORBA is a new generation of distributed computing platform based on the distributed object computing technology, and security is always a fundamental problem of a distributed computing platform. Thus, implementation of a security service in the CORBA platform is very important. Referring to the related security standards and distributed computing platform's security models, the CORBA security service in this paper is designed and implemented conforming to the CORBA Security Service Specification (v1.0). In addition, the necessity of supporting the object invocation monitor in ORB is proposed based on our work on the CORBA security service.

Multi-hop over-the-air reprogramming is essential for the remote installation of software patches and upgrades in wireless sensor networks (WSNs). Recently, coding-based reprogramming protocols are proposed to address efficient code dissemination in environments with high packet loss rate. The problem of analyzing the performance of these protocols, however, has not been explored in the literature. In this paper, we present a high-fidelity analytical model based on Dijkstra’s shortest path algorithm to measure the completion time of codingbased reprogramming protocols. Our model takes into account not only page pipelining and negotiation, but also coding computation. Results from extensive simulations of a representative coding-based reprogramming protocol called Rateless Deluge are in good agreement with the performance predicted by our model, thus validating our approach. Our analytical results show both the number of packets per page and the finite field size have significant impact on completion time. Most notably, the time overhead of coding computation exceeds that of communication when the number of packets per page is 24 and the finite field size is at least 24. Refereed/Peer-reviewed