Design and investigation of scalable multicast recursive protocols for wired and wireless ad hoc networks



D cost = ΣD D cos t ( Lt ),                                                                       (4)

L t T

Where N and L are the total number of routers and the total number of links in the network,
and
Nt and Lt are the number of active routers and the number of active links in the
multicast tree
t, respectively. P cos t and Dcos t denote processing cost at a node and
delivering cost over a link respectively.

Finding the join / leave delay cost should be able to reflect the real network as closely as
possible. In multicast, there are many factors to contribute the join/leave latency, such as how
the registration (towards the source) is carried out, the quantity of multicast traffic needed to
be processed at nodes and over links, the time to build the multicast tree, the size of queue
etc. However, their contributions in occurred latency are varied for different multicast
protocols. For Xcast and Xcast+, no processing of building the multicast tree is needed but
more processing cost at nodes and delivering cost over links, while for SEM the whole
multicast tree is needed to be set up in advance, compared to that for SReM where the
multicast tree is built gradually and could be updated ‘locally’, and less processing and
delivering cost are required for both SEM and SReM. In this thesis for the consideration of
join/leave latency, we assume that the delay cost occurred in the registration processes are
identical for all four protocols. This assumption is reasonable because for multicast protocol
considered, the source is always in charge of tracking the membership of multicast group.
Furthermore, for simplicity, the average processing cost at nodes and average delivering cost
over links are further used to reflect the delay cost, but their contributions to delay cost are
added together. Based on these considerations, join/leave delay cost for a given receiver
i can
be defined as follows:

(5)


Tcost(i)=Ki,nodesPcost+γKi,linksDcost,

where K and K are the number of nodes (routers) and the number of links involved
i,nodes                  i,links

in the receiver i’s join/ leave processing , Pcost and Dcost are given by equation (1) and (2),
respectively.
γ is a weighting factor. Ifγ >> 1, it means the processing cost at nodes plays

85



More intriguing information

1. Optimal Rent Extraction in Pre-Industrial England and France – Default Risk and Monitoring Costs
2. Cardiac Arrhythmia and Geomagnetic Activity
3. The name is absent
4. Der Einfluß der Direktdemokratie auf die Sozialpolitik
5. The name is absent
6. The name is absent
7. The name is absent
8. The name is absent
9. CROSS-COMMODITY PERSPECTIVE ON CONTRACTING: EVIDENCE FROM MISSISSIPPI
10. Human Resource Management Practices and Wage Dispersion in U.S. Establishments
11. Les freins culturels à l'adoption des IFRS en Europe : une analyse du cas français
12. The name is absent
13. Effort and Performance in Public-Policy Contests
14. Do the Largest Firms Grow the Fastest? The Case of U.S. Dairies
15. Anti Microbial Resistance Profile of E. coli isolates From Tropical Free Range Chickens
16. Evolutionary Clustering in Indonesian Ethnic Textile Motifs
17. The name is absent
18. The Evolution
19. Nonparametric cointegration analysis
20. The name is absent