Scalability improvement of the real time control protocol

El-Marakby, R.; Hutchison, David

doi:10.1016/j.comcom.2004.06.006

Cited by 3 publications

(3 citation statements)

References 16 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Among the solutions proposed for RTCP's incompetencies, S-RTCP [2][3][4] has proved to be able to solve many of the problems associated with RTCP. However, S-RTCP introduces new problems on its own.…”

Section: Introductionmentioning

confidence: 99%

“…This study makes an effort in improving S-RTCP by proposing series of reformations and new features, in terms of stability and packet loss. Introduction to S-RTCP: S-RTCP applies a multilevel hierarchical architecture to RTP session in order to arrange RTP session participants and their reports [2][3][4] . S-RTCP uses hierarchy to restrain RTCP RR reports in smaller groups, as Fig.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Error Resistant Real-Time Transport Control Protocol

Shahbazi¹,

Jumari²,

Ismail³

2009

American J. of Engineering and Applied Sciences

View full text Add to dashboard Cite

Problem statement: Real-time Transport Control Protocol (RTCP) protocol has been the subject of various criticisms due its problematic performance in large-scale networks. S-RTCP is a protocol with high potential as it has proved to be able to solve many problems of RTCP. It has numerous flaws on its own. This study aimed at dealing with flaws of S-RTCP and improving it in terms of stability and packet loss. Approach: A new proposed scheme was designed. Modifications included designing multi-manager scheme, improving parent-seeking procedures, reducing distribution of request packets, reforming the design to be independent from TTL, adding methods to check on sanity of manager nodes. This study considered packet loss ratio of below 2% as desirable. Results: ER-RTCP comparing to legacy RTCP in terms of packet loss using NS-2 in four different scenarios revealed improvements between 73 and 88% for various scenarios. It also kept packet loss rate below 2% for all scenarios. Comparison of ER-RTCP to S-RTCP showed that based on different α (stability of each single manager) values, ER-RTCP was more stable as it showed more resistance to entire scheme breakdown (β). ER-RTCP's parent-seeking procedure, as modeled scenario revealed a packet generation reduction of 97%, compared to S-RTCP's. In occurrence of parent AG leave or loss, ER-RTCP reduced request packet generation by 95%. Allowance of AG dismissing in ER-RTCP, avoided occurrence of packet loss, as sample scenario showed S-RTCP experiencing packet loss of 3.5% while ER-RTCP kept packet loss at zero in theory. Conclusion: Proposed design improved S-RTCP in terms of reduction of packet loss and stability.

show abstract

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Error Resistant Real-Time Transport Control Protocol

Shahbazi¹,

Jumari²,

Ismail³

2009

American J. of Engineering and Applied Sciences

View full text Add to dashboard Cite

show abstract

“…characteristics of wired and wireless network nodes has several applications including localization of mobile Region formation based on physical distances has nodes [3], enhancing the scalability and QoS control several advantages as it allows controlling the size and [8], and supporting the use of region-based spatial shape of the regions. However, the physical location of logic to reason about networked systems [4].…”

mentioning

confidence: 99%

Spatial Clustering of Wireless Sensor Network Nodes Based on Packet Loss

Tawfik

El-Marakby

2007

2007 IEEE International Symposium on Signal Processing and Information Technology

Self Cite

View full text Add to dashboard Cite

This paper focuses on region formation in Wireless This paper examines the problem ofgrouping wireless Sensor Networks (WSNs) as region formation plays an sensor nodes forming a network into a set of spatial important role in facilitating the programmability clusters. The technique for grouping the nodes [14,15], the scalability [10], collision reduction, and proposed here, exploits the spatial degradation in routing [1,18] in WSNs. Spatial aspects are also useful wireless signals to group the nodes into regions based in query processing [12], and for collaborative medium on the packet loss rate. The technique is validated on access control [13]. The term "spatial clustering" datasets obtained from actual wireless sensor denotes the formation of spatial regions that include a networks. The obtained results support the validity of set of neighboring sensors. Typically, a cluster consists the proposed approach. of sensors grouped according to a shared property. Shared properties can be a communication feature such 1. Introduction as the number of hops between any pair of sensors within the cluster [17], the actual distances separating sensors [13], or a sensing property like the correlation Forming spatial regions based on the communication between sensor readings [14]. characteristics of wired and wireless network nodes has several applications including localization of mobile Region formation based on physical distances has nodes [3], enhancing the scalability and QoS control several advantages as it allows controlling the size and [8], and supporting the use of region-based spatial shape of the regions. However, the physical location of logic to reason about networked systems [4].sensors is not always known. In many applications, sensors are randomly deployed and in others the sensor In general, a region is a contiguous finite extent of may drift under the effect of environmental factors like space surrounded by a boundary. Therefore, a region wind, rain, etc. Spatial clustering of sensors based on divides the space into three elements: a boundary, a communication properties allows us to form groups of finite internal space, and a possibly infinite external sensors that are physically close but does not ensure space. This conception of a region applies to one the size, shape, or even the continuity of the resulting dimensional, two dimensional, and three dimensional regions. However, there are advantages of this method spaces [7]. To apply this definition to network nodes, of clustering sensors. First, we can control the number the nodes must be grouped into localities according to of the resulting clusters. Moreover, we can some measurable quantity related to distance. For dynamically re-cluster the sensors as they drift or example, the distance can be represented by the hop move. Using sensing properties for clustering sensors count separating network nodes, or even better, the is highly dependent on the spatial properties of the Euclidean distances, if the location of each network physical phenomenon being se...

show abstract