Providing recall-guaranteed search is critical for P2P networks. While building semantic overlay improves search performance, existing designs suffer from a tradeoff between search time and search quality (i.e. high recall). Moreover, they require to use high control overhead for overlay maintenance. In this paper, we present rSearch to achieve fast search with guaranteed high recall. The rSearch-enabled overlay topology looks like a ring, augmented with semantic chord links. Given a query, rSearch uses multiple query walkers that traverse on the ring independently to find relevant semantic nodes for answers. The ring structure facilitates fast and low-redundancy query forwarding, while the abundant semantic chord links enable large semantic clusters. Bloom Filter is used to encode and compress node semantic summaries, greatly saving control overhead. rSearch further considers churn resilience and network awareness to enhance system performance. Extensive simulations with real-life file sharing trace and network latency trace show that rSearch greatly outperforms GES.
{"title":"rSearch: Ring-Based Semantic Overlay for Efficient Recall-Guaranteed Search in P2P Networks","authors":"Zhenyu Li, Gaogang Xie","doi":"10.1109/ICPPW.2010.55","DOIUrl":"https://doi.org/10.1109/ICPPW.2010.55","url":null,"abstract":"Providing recall-guaranteed search is critical for P2P networks. While building semantic overlay improves search performance, existing designs suffer from a tradeoff between search time and search quality (i.e. high recall). Moreover, they require to use high control overhead for overlay maintenance. In this paper, we present rSearch to achieve fast search with guaranteed high recall. The rSearch-enabled overlay topology looks like a ring, augmented with semantic chord links. Given a query, rSearch uses multiple query walkers that traverse on the ring independently to find relevant semantic nodes for answers. The ring structure facilitates fast and low-redundancy query forwarding, while the abundant semantic chord links enable large semantic clusters. Bloom Filter is used to encode and compress node semantic summaries, greatly saving control overhead. rSearch further considers churn resilience and network awareness to enhance system performance. Extensive simulations with real-life file sharing trace and network latency trace show that rSearch greatly outperforms GES.","PeriodicalId":415472,"journal":{"name":"2010 39th International Conference on Parallel Processing Workshops","volume":"8 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2010-09-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"125271778","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
In wireless sensor network, sensory readings are often noisy due to the imprecision of measuring hardware and the disturbance of deployment environment, so it is often inaccurate if we use individual sensor readings to answer queries. In this paper, we consider a useful application of sensor network: maximum average value region query. This query returns the region with the maximum average value among all possible regions in the network, where the region is a fix-sized circle pre-defined by users. Using the average value of a region to answer the query, noises between sensors will be neutralized with each other, which will make the results more reliable. However, because of the huge amount of possible regions in the network, it is costly to process the query exactly. Therefore, we propose a sampling-based algorithm AMAVR to deal with the problem approximately. AMAVR uses a background value to prune the useless regions which cannot be the result. A further optimization strategy is also given to handle the situation that, background value based filter does not work when some individual sensor nodes have higher values than their neighbors. By using both of the two techniques, the scale of the sampling population can be effectively reduced, that is, we cost less energy to get a satisfying result. At last, the conducted simulations demonstrate the energy efficiency of the proposed methods in our paper.
{"title":"A Sampling-Based Algorithm for Approximating Maximum Average Value Region in Wireless Sensor Network","authors":"H. Zhang, Zhongbo Wu, Deying Li, Hong Chen","doi":"10.1109/ICPPW.2010.14","DOIUrl":"https://doi.org/10.1109/ICPPW.2010.14","url":null,"abstract":"In wireless sensor network, sensory readings are often noisy due to the imprecision of measuring hardware and the disturbance of deployment environment, so it is often inaccurate if we use individual sensor readings to answer queries. In this paper, we consider a useful application of sensor network: maximum average value region query. This query returns the region with the maximum average value among all possible regions in the network, where the region is a fix-sized circle pre-defined by users. Using the average value of a region to answer the query, noises between sensors will be neutralized with each other, which will make the results more reliable. However, because of the huge amount of possible regions in the network, it is costly to process the query exactly. Therefore, we propose a sampling-based algorithm AMAVR to deal with the problem approximately. AMAVR uses a background value to prune the useless regions which cannot be the result. A further optimization strategy is also given to handle the situation that, background value based filter does not work when some individual sensor nodes have higher values than their neighbors. By using both of the two techniques, the scale of the sampling population can be effectively reduced, that is, we cost less energy to get a satisfying result. At last, the conducted simulations demonstrate the energy efficiency of the proposed methods in our paper.","PeriodicalId":415472,"journal":{"name":"2010 39th International Conference on Parallel Processing Workshops","volume":"56 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2010-09-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"128475672","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Balancing fairness, user performance, and system performance is a critical concern when developing and installing parallel schedulers. Sandia uses a customized scheduler to manage many of their parallel machines. A primary function of the scheduler is to ensure that the machines have good utilization and that users are treated in a ``fair'' manner. A separate compute process allocator (CPA) ensures that the jobs on the machines are not too fragmented in order to maximize throughput. Until recently, there has been no established technique to measure the fairness of parallel job schedulers. This paper introduces a ``hybrid'' fairness metric that is similar to recently proposed metrics. The metric uses the Sandia version of a ``fairshare'' queuing priority as the basis for fairness. The hybrid fairness metric is used to evaluate a Sandia workload. Using these results, multiple scheduling strategies are introduced to improve performance while satisfying user and system constraints.
{"title":"Parallel Job Scheduling Policies to Improve Fairness: A Case Study","authors":"V. Leung, Gerald Sabin, P. Sadayappan","doi":"10.1109/ICPPW.2010.48","DOIUrl":"https://doi.org/10.1109/ICPPW.2010.48","url":null,"abstract":"Balancing fairness, user performance, and system performance is a critical concern when developing and installing parallel schedulers. Sandia uses a customized scheduler to manage many of their parallel machines. A primary function of the scheduler is to ensure that the machines have good utilization and that users are treated in a ``fair'' manner. A separate compute process allocator (CPA) ensures that the jobs on the machines are not too fragmented in order to maximize throughput. Until recently, there has been no established technique to measure the fairness of parallel job schedulers. This paper introduces a ``hybrid'' fairness metric that is similar to recently proposed metrics. The metric uses the Sandia version of a ``fairshare'' queuing priority as the basis for fairness. The hybrid fairness metric is used to evaluate a Sandia workload. Using these results, multiple scheduling strategies are introduced to improve performance while satisfying user and system constraints.","PeriodicalId":415472,"journal":{"name":"2010 39th International Conference on Parallel Processing Workshops","volume":"25 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2006-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"121852209","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：