An Adaptive Reliable Multipath Centralized RWA Algorithm for All-Optical WDM Networks

 The abstract of this paper is :

Balancing the load among multiple paths leads to the degradation in the all-optical WDM network. It is always desirable to deliver the entire load on a single optimal path depending upon the current network-wide load status. For this case, multiple light paths need to be maintained to provide the choice of selecting the best light path, based on the changing traffic load conditions, This paper presents the design and analysis of an adaptive reliable multipath centralized routing algorithm (ARMCR) with two central management systems (CMS-1 and CMS-2) which considers load heuristic for selecting the optimal and backup paths. Initially in this algorithm when a request comes to a source it sends the information to CMS-1 for allocating the optimal primary and backup paths for the received request, depending on the number of available free wavelengths. With the assigned primary path, source starts clearing the request meanwhile CMS-2 keeps track of the changes in the available free wavelengths and failure of all the paths. In the assigned primary path if the available free wavelength number comes below the threshold number ‗N‘ and at the same time if any other path is available with more number of free wavelengths, then CMS-2 assigns this path as primary path and very next more number available free wavelengths path is assigned as backup path. The advantage of this algorithm is that it reduces the set up time by using two CMS which share the work and also if one fails the other will take over the entire process. As opposed to the reactive protocols, our proposed protocol is proactive in the sense that it avoids the chances of blocking. Furthermore, the approach is self-regulating, it automatically adapts to the traffic load variation across the network. By simulation results, we showed that our proposed protocol has low set up time and blocking.

Full paper you can read and download from the following link, 

http://www.ciitresearch.org/dl/index.php/nce/article/view/NCE042011008

If you have any doubts in this paper feel free to contact JARVIS HELP team.

Written By:
SANDY -- (Jarvishelp Member)

For more information:
www.jarvishelp.in
mail@jarvishelp.in

IP Address Classes

The communication between two devices over a particular medium wired/wireless is called networking. To communicate in a wired network every system must and should have a valid IP address. IP address in an identification tag or key for a system. IP address looks like 192.168.10.21. Based on these IP address, they can be classified into five types; Class A, Class B,Class C, Class D and Class E. 

Each IP address can be divided into four octettes, IP address can be written as Octet1.Octet2.Octet3.Octet4. Here Octet 1 represents network class type. 

 Class A Network:

Class-A address are used for networks with very higher number of systems. Range for Octet-1 value of Class-A network is 1-126. Default subnet mask for Class-A type is 255.0.0.0. Number of possible network ID's are 126.

127 is excluded from Class-A, Ranges 127.X.X.X are reserved for the loopback or localhost, example 127.0.0.1 is the common local host address. 

 Class B Network:

Class-B addresses are used for medium-sized to large-sized networks. Range for Octet-1 value of Class-B network is 128-191. Default subnet mask for Class-B type is 255.255.0.0. Number of possible network ID's are 16,384.

 Class C Network:

Class-C address are used for smaller networks, like university campus. Range for Octet-1 value of Class-C network is 192-223. Default subnet mask for Class-A type is 255.255.255.0. Number of possible network ID's are 20,97,152.

 Class D Network:

Class-D network addresses are reserved for multicast purpose. Range for Octet-1 value of Class-D network is 224-239.

 Class E Network:

Class-E network addresses are reserved for Future purpose. Range for Octet-1 value of Class-E network is 240-255.

Range 255.255.255.255 broadcasts to all hosts on the local network.

Written By:
SANDY -- (Jarvishelp Member)

For more information:
www.jarvishelp.in
mail@jarvishelp.in

Why RWA is a challenge in all-optical WDM networks?

How the data requests will be served in Optical - WDM Networks ?

In Optical WDM Networks, when ever a data request arrives at a Node, first it validates the request, during this process all the headers in the request will be validated, after validating request it identifies the source node and destination node addresses. Once it get the Source-Destination pair it will assign a route and wavelength to transmit. This assignment is valid till all the data is transmitted. 

What is RWA?

RWA -- "Routing and Wavelength Assignment", As described above once a node identifies source and destination pair the process of assigning a route and wavelength is called as RWA Protocol. There are many RWA protocols exists and depends on the requirement and network type the protocol will be choose or implemented.

What are the main network types in WDM Networks?

 

 Mainly there two types of Network architecture but these days hybrid implementation also being used so we can say three types. They are Distributed Networks, Centralized Networks and Hybrid or Global Networks.

In Distributed Networks each and every node in network will have the information of entire network structure, in future if there is any new node added or removed that information need to be updated at all the nodes in network. Along with this each node maintain a minimum information of paths between source and destination pair. It also keep track of wavelengths assigned.

In  Centralized Networks all the network information will be stored at Central Management System and every request will be processed or served by CMS. In this case if there is any modification in network then it is enough if we update only at CMS. But major dis advantage is when there is failure in CMS the entire network will be down. In order to avoid this in the design people will maintain the backup CMS also. 

In hybrid or Global Networks, they will use the combination of both Distributed and Centralized Network types, means small small groups will in Centralized manner and all these CMS will be a part or act as nodes of a Distribute Network. The major advantage of this is if there is any failure in network a particular group or very few number of devices will be affected.

Why RWA is challenge?

There are many routing and wavelength assignment techniques are available and depends on the technique the cost, delay, resource utilization and QoS are effected. If we use standard technique like shortest path algorithm then there is more chance of congestion in the network, If use the Wavelength Continuity Constraint WCC algorithm then cost of request can be reduced but the number of calls or requests served will be limited. If we choose a path based on blocking of calls then there is a chance of choosing a long path though which latency increases. Like this many other techniques are available for RWA so based on our requirement we need to select which algorithm is suitable like shortest path, Less blocking path. Wavelength also need to be choosed based on requirement like least used, most used, WCC or some other. So ultimately RWA plays an important role in network Quality of Service. The major contributors are the number of nodes on path, number of wavelength conversions required and number of wavelengths are utilized on links.

Written By:
SANDY -- (Jarvishelp Member)

For more information:
www.jarvishelp.in
mail@jarvishelp.in