A scalable content-addressable network

RatnasamySylvia,; FrancisPaul,; HandleyMark,; KarpRichard,; Shenker, Scott

doi:10.1145/964723.383072

Cited by 2,333 publications

(2,057 citation statements)

References 12 publications

Supporting

Mentioning

2,001

Contrasting

Unclassified

Order By: Relevance

“…DCS uses the GPSR [7] geographic routing algorithm for low-level routing. It then builds a DHT [11,12,14,16] on top of GPSR. We first describe GPSR and then the construction of a lookup algorithm over GPSR.…”

Section: Data-centric Storagementioning

confidence: 99%

“…A significant benefit of DCS is that queries for data of a certain type can be sent directly to the node storing data of that type, rather than flooding the queries throughout the network (unlike data-centric routing proposals [4,6]). DCS is based on the low-level routing functionality provided by the GPSR geographic routing algorithm [7], and on distributed hash-table functionality provided by peer-to-peer lookup algorithms such as CAN [11], Chord [14], Pastry [12] and Tapestry [16]. It is shown in [10] and [13] that DCS offers reduced total network load and peak (hotspot) network usage.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Resilient Data-Centric Storage in Wireless Ad-Hoc Sensor Networks

Ghose

Großklags

Chuang

2002

Mobile Data Management

View full text Add to dashboard Cite

Abstract. Wireless sensor networks will be used in a wide range of challenging applications where numerous sensor nodes are linked to monitor and report distributed event occurrences. In contrast to traditional communication networks, the single major resource constraint in sensor networks is power, due to the limited battery life of sensor devices. It has been shown that data-centric methodologies can be used to solve this problem efficiently. In data-centric storage, a recently proposed data dissemination framework, all event data is stored by type at designated nodes in the network and can later be retrieved by distributed mobile access points in the network. In this paper we propose Resilient Data-Centric Storage (R-DCS) as a method to achieve scalability and resilience by replicating data at strategic locations in the sensor network. Through analytical results and simulations, we show that this scheme leads to significant energy savings in reasonably large-sized networks and scales well with increasing node-density and query rate. We also show that R-DCS realizes graceful performance degradation in the presence of clustered as well as isolated node failures, hence making the sensornet data robust.

show abstract

Section: Data-centric Storagementioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Resilient Data-Centric Storage in Wireless Ad-Hoc Sensor Networks

Ghose

Großklags

Chuang

2002

Mobile Data Management

View full text Add to dashboard Cite

show abstract

“…The edges in U may be unidirectional or not. We believe that DHT graphs [3,4,5] can be efficiently used as the graph (X, U). In this paper, we use the Chord DHT graph as an example.…”

Section: Topology-aware Multi-cluster Overlaymentioning

confidence: 99%

“…The floods support arbitrary queries, but are not scalable because they cause exponentially increased network traffic. In contrast, structured P2P systems are developed to perform key queries by constructing Distributed Hash Tables (DHTs), such as Chord [3], CAN [4], and Pastry [5], etc. Although such schemes provide good performance for exact match queries, they almost don't work for range, approximate, or text queries.…”

Section: Introductionmentioning

confidence: 99%

Topology-Aware Multi-cluster Architecture Based on Efficient Index Techniques

Zhao

Zhang

et al. 2005

Lecture Notes in Computer Science

View full text Add to dashboard Cite

Abstract. In this paper, we focus on how to construct an efficient unstructured P2P system. The main contributions of our proposal are two-fold. First, aiming at alleviating the topology mismatch problem between the P2P logical overlay network and the physical underlying network, we proposed a Topology-aware Multi-cluster Overlay (TMO) architecture where peers self-organize into clusters based on network locality. Second, in order to further improve the search efficiency of the TMO architecture, we present two novel index techniques, namely, cluster-index technique and topic-index technique. The two different techniques are highly effective in different application domains in which the TMO architecture is deployed. The simulation results indicate that our proposed schemes are efficient in both resource usage and data retrieval.

show abstract

“…Overlay networks such as the Content Addressable Network (CAN), 2 Chord, 3 Pastry, 4 and Viceroy 5 create a virtual topology on top of the physical topology. In this sense, TTL-based P2P networks are also a type of overlay, but we use the term here to refer only to networks that create virtual topologies based on node-content attributes.…”

Section: Overlay Networkmentioning

confidence: 99%