A distributed file system (DFS) is a client/server-based application that allows clients to access and process data stored on the server as if it were on their own computer. An edge router has series of data cards, to handle data inflow and outflow which can be of the form of email headers, data packets, etc. Each of these data cards have their own file system architecture. This paper implements a distributed file system approach on all these data cards, so as to make it a centrally controlled one file system and not having parallel many file systems. Convenience of usage and file handling capacity is also looked in this paper. The servers are portrayed as the data cards of an edge router and the client access points are the router processor cards. The DFS on Edge Router is implemented using GlusterFS which is a scale-out network-attached storage file system.
IP address lookup to route data packets is an important function in a router and improving this function improves the overall performance of the router. From the data structures used for the prefixes representation, there are the trees that represent prefixes with their binary values. However, this data structure requires an improvement because of the longest prefix function search complexity. Our approach is used to improve the routing information search time in the prefixes values binary tree by periodically reorganizing the tree according to the use of prefixes; the most recently used prefixes are stored in the higher levels of the reorganized binary prefixes value tree (RBPVT) which improves the data packets routing time. The tests and evaluation of the access memory number of the longest prefix match search algorithm shows that the IP address lookup algorithm based on our RBPVT tree improves the performance of the IP routers in terms of average memory access number.
Cognitive Radio (CR) is an intelligent technology used to sense the spectrum in order to make efficient use of limited spectrum available. The use of such radios in ad hoc network not only increases the probability of reach ability but also provides a reliable path between a pair of nodes. Networks that employ CR technology are termed as Cognitive Radio Ad Hoc Networks (CRAHN). Quality of route depends upon several parameters but it is highly influenced by end to end delay encountered by a packet to reach from source to destination. More is the end to end delay; lesser will be the efficiency of the network. End to end delay is directly concerned with the reliability of data transfer. Therefore this paper proposes a routing scheme that obtains a most reliable route with minimum end to end delay using fuzzy logic theory. The proposed strategy is implemented in MATLAB-7.0 and the simulation results show that our proposed routing scheme outperforms the traditional Shortest Spectrum Aware routing scheme available in literature.
Today
Yesterday
This week
This month
All days
Number of views: 336