Through analysis of the process of handoff and existing improvement program, a novel Cache update solution is proposed. It is composed of slice Cache update algorithm, which contains a Cache update trigger mechanism b...
详细信息
Through analysis of the process of handoff and existing improvement program, a novel Cache update solution is proposed. It is composed of slice Cache update algorithm, which contains a Cache update trigger mechanism based on dynamic domain values, AP(access point) load balancing and a Cache update strategy based on slice channel scanning. Experiment results show that this solution can effectively avoid Cache updates when STA (station) is stable and the associated AP's signal is good. While ensuring the timely update of the Cache it can reduce the cost of each Cache update and the handoff delay. Meanwhile the load balancing of each AP can be well considered while choosing the AP. Therefore, the solution is able to significantly improve the mobile terminals handoff performance between AP.
Due to the deficiency of fluid-based (FB) P2P network model in dealing with the stream which could not be divided infinitely, a limitedly dividable stream-based (LDSB) model was studied and proposed for P2P streaming ...
详细信息
Due to the deficiency of fluid-based (FB) P2P network model in dealing with the stream which could not be divided infinitely, a limitedly dividable stream-based (LDSB) model was studied and proposed for P2P streaming networks. A close-form expression of maximum universal streaming rate of LDSB model was given and proved in theory. The results of theoretical analysis and numerical calculation show that the maximum universal streaming rate of LDSB model can not be greater than that of FB model because of the limitation of minimum sub streaming rate. However, with the decrease of minimum sub streaming rate, the performance of LDSB model would approximate to FB model. In addition, by choosing appropriate system parameters such as minimum sub streaming rate and server upload rate, etc., the performance of LDSB system may reach the upper limit of FB model's theoretical performance, and unnecessary waste of resources in system deployment can be avoided.
Component-based development is gradually showing its advantages in building complex systems with shorter time and less cost than traditional methods. However, mismatching and semantic are key problems in component sea...
Component-based development is gradually showing its advantages in building complex systems with shorter time and less cost than traditional methods. However, mismatching and semantic are key problems in component searching process. In this paper, a function description model is proposed to precisely and completely describe the user requirements and component function based on domain models. In the component selection process, a sophisticated matching algorithm is introduced for semantic problems in matching two activity profiles derived from different domain models. A component selection method is also presented to improve interoperability for multi domain models, followed by the implemented prototype of the proposed methods.
Node churn is one hurdle to using peer-to-peer (P2P) networks as anonymous networks, which makes the anonymous path fragile and results in message loss and communication failures. A collaboration scheme including frie...
详细信息
ISBN:
(纸本)9781424416509
Node churn is one hurdle to using peer-to-peer (P2P) networks as anonymous networks, which makes the anonymous path fragile and results in message loss and communication failures. A collaboration scheme including friendly neighbor-based incentive (FNI) and re-encryption mechanism is proposed to deal with the high node churn (changes in system membership) characteristic of unstructured P2P networks. The simple FNI mechanism is presented to encourage peers to forward other peers' queries, and establish more connections to improve the performance of P2P overlay network where only stable and well-behaved nodes can be chosen as relay nodes to prolong single path durability. The re-encryption mechanism is designed to replace those failed relay nodes and achieve routing resilience upon different node availabilities in real- world systems. The results from our security analysis and simulation show that the collaboration scheme greatly improves routing resilience and maintains low latencies and low communication overhead.
Although a large number of different proposals have emerged in recent years, a fundamental problem of P2P streaming networks concerned about is to establish an optimal routing structure with achieving certain optimiza...
详细信息
ISBN:
(纸本)9781424432806
Although a large number of different proposals have emerged in recent years, a fundamental problem of P2P streaming networks concerned about is to establish an optimal routing structure with achieving certain optimization metric. In this paper, we focus on the problem of maximizing universal streaming rate in P2P streaming networks with constrained neighborhood. Firstly, we extend the fluid-based model and propose a new P2P streaming network model taking the neighborhood constraint into account. Second, we formulate the problem of maximizing the universal streaming rate in the new model. After proving its NP-hardness, we propose a heuristic multiple trees packing algorithm and corresponding optimum strategies. The simulation results show that the algorithm can achieve the theoretical upper bound of the universal streaming rate in fluid-based model without constrained neighborhood. While neighborhood constraint is considered, in addition, the universal streaming rate will have different decline according to various intensity of neighborhood constraint.
Multi-objective flowshop scheduling problems have gained wide attention both in practical and academic fields. In this paper, a hybrid multi-objective genetic algorithm is proposed to solve multi-objective no-wait flo...
详细信息
Multi-objective flowshop scheduling problems have gained wide attention both in practical and academic fields. In this paper, a hybrid multi-objective genetic algorithm is proposed to solve multi-objective no-wait flowshop scheduling problems with both the makespan and the total flow time minimization. The proposed algorithm makes use of the principle of non-dominated sorting, coupled with the use of elitist solutions strategy. To improve efficiency, two effective multi-objective local search procedures based on objective increment are applied on each chromosome in the archive. Experimental results show that the proposal outperforms the other three heuristics both on effectiveness and efficiency.
No-wait flowshops with flowtime minimization are typical NP-Complete combinatorial optimization problems, widely existing in practical manufacturing systems. Different from traditional methods by which objective of a ...
详细信息
No-wait flowshops with flowtime minimization are typical NP-Complete combinatorial optimization problems, widely existing in practical manufacturing systems. Different from traditional methods by which objective of a new schedule being completely computed objective increment methods are presented in this paper by which the objective of an offspring being obtained just by objective increments and computational time can be considerably reduced. HGAI (Hybrid GA based on objective Increment) is proposed by integrating genetic algorithm with a local search method. A heuristic is constructed to generate an individual of initial population and a crossover operator is introduced for mating process. HGAI is compared with two best so far algorithms for the considered problem on 110 benchmark instances. Computational results show that HGAI outperforms the existing two in effectiveness with a little more computation time.
In this paper, no-wait flow shop scheduling problem with total flowtime minimization is considered. A hybrid heuristic is proposed, which is based on PH1 (p) (presented by Aldowaisan and Allahverdi, OMEGA, 2004). A co...
详细信息
The Eleman Neural network has been widely used in various fields ranging from temporal version of the Exclusive-OR function to the discovery of syntactic categories in natural language date. However, one of the proble...
详细信息
In this paper, no-wait flow shop scheduling problem with total flowtime minimization is considered. A hybrid heuristic is proposed, which is based on PHI (p) (presented by Aldowaisan and Allahverdi, OMEGA, 2004). A co...
详细信息
ISBN:
(纸本)1424409624
In this paper, no-wait flow shop scheduling problem with total flowtime minimization is considered. A hybrid heuristic is proposed, which is based on PHI (p) (presented by Aldowaisan and Allahverdi, OMEGA, 2004). A composite algorithm is adopted to generate the initial seed. Job insertion in PHI (p) is replaced with an existing constructive heuristic. Experimental results show that the proposal outperforms PHI (p), especially for large scale instances.
暂无评论