A three Stage Load Sharing Routing Algorithm to Increase lifetime of Cognitive Radio Sensor Networks



Download 38.98 Kb.
Date05.08.2017
Size38.98 Kb.
#26778

A Three Stage Load Sharing Routing Algorithm to Increase lifetime of Cognitive Radio Sensor Networks


Arslan Haider , Arslan Ali

Government College University Lahore, Pakistan

{ arslanhaiderranjha1@gmail.com; arslan_ali_pak@yahoo.com}



Abstract— A Cognitive Radio Sensor Network (CRSN) is a network of wireless cognitive radio sensor nodes, which sense data and collaboratively transmit over dynamically available spectrum bands in a multi-hop manner. The power of these nodes reduces during transmission and eventually the system dies. Traditional routing algorithms are not suitable for CRSN because these algorithms are based on the shortest hops whereas in cognitive radio environment the status of each link is different. In this paper we introduce a three-stage routing algorithm for efficient performance. In the first stage we calculate the transmission delay (in form of link cost) for every link, in the second stage cost of each link is adjusted between primary radio sensor nodes and cognitive radio sensor nodes according to the state of primary users and in the third stage an efficient load sharing routing technique is applied to increase the life time of all the sensor nodes in a way to get maximum efficiency from the network. We modify the traditional routing method by saving the power of each node. Computer simulation results clearly indicate an increase in lifetime of a network.
Keywords: Cognitive radio sensor network; Energy efficiency; Efficient load sharing routing algorithm.

Introduction


Cognitive radio is an emerging wireless communications concept in which a network or a wireless node is able to sense its environment, especially spectrum holes, and change its transmission and reception chains to communicate in an opportunistic manner, without interfering with licensed users. Cognitive radio thus aims to improve the way the scarce radio spectrum is utilized [1][2]. Cognitive radio (CR) has received increasing attention in the past few years as a promising technology to implement the dynamic spectrum access and improve the spectral efficiency [3]-[5]. The usage of CR networks has been recognized in many areas like to disaster rescue after the earthquake, public safety monitoring [6], for tactical communications in a battlefield, military sensing and tracking [7], etc.

Routing in multi-hop CR networks presents a great challenge mainly due to the following two facts. First, the traditional wireless links noticeably degrade the reception quality of primary radios and these are not reliable because of channel fading [8]-[10]. Second, a CR node immediately interrupts its transmission whenever a neighboring PR activity is detected. This requires frequent monitoring of the PR activities [11]. Nodes are the building blocks of (CRSNs) and are very low cost, low power computers that can monitors one or more sensors. These nodes are built up of a small battery, sensors and memory elements. These sensor nodes have limited range of receiving and transmitting data. In order to specify the path for the data transmission, these nodes have to make a mutual coordination with its nearest neighbor. These nodes behave like a router for transceiving information. Power supplied to an individual node completely depends upon the small battery of limited energy inside it. Placements of nodes in improper places and difficulty in charging or recharging batteries have made researchers to do investigations on reduction of energy consumption [12]. For increasing lifetime, a route is selected for data transmission in which the number of errors and conflicts is low and hence the number of sending messages decreases in the network and network lifetime increases [13]. Generally routing algorithms like Link State choose the best route in term of low cost. It does not include the energy statistics of each sensor. It transmits data from one node to another in lowest possible time but some specific nodes are used more frequently than others which cause such nodes to run short of power resources and decreases the lifetime of a network.

In this paper, we firstly define cost that takes the characteristics of wireless links into account and then an efficient load sharing routing technique is applied that not only selects the best route in terms of cost but also increase lifetime of a network by managing the power resource of each node in Cognitive Radio Sensor Network.

The remainder of the paper is organized as follows. In Section II, we describes the limitations of traditional routing algorithms. In Section III we present the delay estimation technique and the efficient load sharing routing algorithm. Transmission behavior of proposed algorithm is elaborates in Section ІV. Finally, section V summarizes the remarks in the conclusion.



  1. Simple CR sensor network with Primary radio node

Limitations of traditional routing algorithms


Traditional Routing algorithm for CR networks does not always give the best solution in terms of power efficiency. Consider a simple Cognitive Radio Sensor Network as shown in Figure 1. There are a number of routes from source node A to destination node F. A has two neighbors (B and C) and all the packets for destination F are forwarded by one of these two neighbors. Traditional routing is not suitable to CRSNs because in cognitive radio environment, the status of every link is different whereas the traditional routing techniques are based on the shortest hops. Applying traditional routing algorithm on the network shown in Fig 1 will give route A-B-D-E. However, the links in Fig.1 have different channels bandwidth and different probabilities, therefore the traditional routing algorithm do not consider these type of factors, so the cost of route A-B-D-E chosen by traditional routing algorithm may not be the most appropriate.

Traditional routing does not take into account the delivery rate of CR links, i.e. the dashed links. For example when the primary user is “OFF”, the CR link is available for packet delivery but when the primary user is “ON”, the delivery rate of the CR link should be zero so that the CR users cannot cause interference to the licensed users[5]. The CR link can be considered as a two state link. For example when the primary user is “OFF” the delivery rate of link C-E-F will be maximum (due to low link cost adjusted by proposed algorithm) whereas when the primary user is “ON” the delivery rate of link C-E-F will be 0. (Due to very high link cost adjusted by proposed algorithm). In the absence of primary users the packet is relayed to node F through node D and delivery fails only when both links between C and F are broken. Traditional Algorithms are used to define a low cost path for the transmission of packets. A CRSN is made of small low power sensors or nodes.

One of the major problem in Traditional algorithms is that it does not take into account energy resources of a network. It always try to find a minimum cost routing path to reduce the transmission time. This sometime results in an unjustified distribution of load. Some nodes which are associated with a low-cost path may be distributed with heavy loads as compared to other nodes. This results in a rapid consumption of energy resource for such nodes. A node is considered as dead when it consumes all of its available energy and all the nodes connected to network through this node get into isolation. This seriously effects the lifetime of overall network.

EFFICIENT ROUTING ALGORITHM FOR CRSN


The proposed design overcome these issues by calculating cost and then using a modified link state algorithm. In this section, we present our novel and energy efficient cognitive routing Algorithm for CRSNs that find a path from CR source node to CR destination node.

Network model

Before presenting the algorithm, we will describe the network model and the assumptions. We consider number of nodes in network. There is a set of primary radio nodes denoted by and a set of Cognitive Radio nodes (CR) denoted by. Let be any cognitive radio node, be its neighbor cognitive radio node and is primary radio node. First step is that link cost is calculated between any two nodes and. There are M number of channels are available between two nodes. Each channel has its available probability at every link. The bandwidth of channel between node and is and the available probability of channel is. Let the active probability of primary radio nodes in channel is . So CR nodes can utilize the channel with the available probability. The maximum channel capacity (is given by Shannon’s Theorem:


(1)
Where is signal to noise ratio of channel between node and node. We define effective link capacity between node and nodeas follows:

=.) (2)

If we given the length of packets as L, then we can calculate the transmission delay as follows:


(3)
We then define cost of the link between node and node as:

= (4)
Similarly
= (5)
After finding the cost of each link, the algorithm adjusts the cost of links between primary radio sensor nodes and cognitive radio sensor nodes according to the state of primary users. When the primary user is “ON” it increases the costs of all the links which are connected to primary radio node to a very high value and the load at this node will be routed through some other alternative paths. On the other hand it does not affect link costs when the primary user is “OFF”. Then the modified algorithm will be applied that chooses the low cost path for data transmission using link state algorithm and also saves the power of a node by sharing load between different nodes.

Figure 2. A small network to examine behavior of a critical node


The proposed design works by continuous monitoring. Each node monitors its power continuously. We declare a threshold value of power for a network. If the value of power for all the nodes involved is greater than threshold value, then a path is chosen for transmission. Any node with power less than threshold value will be declared as critical node. A critical node advertises to its neighbors about the current state and increases the cost of all the connected links to a very high value therefore the load at this node will be routed through some other alternative paths even if it is slightly higher in cost. A node with power less then threshold power will only be used if there is no other substitute path available for transmission. This technique not only increases the transmission rate but a significant increase in lifetime of a CR sensor network is also achieved. This algorithm significantly reduces the load on critical node and increases the life a cognitive radio sensor network. The algorithm is given below

Proposed Routing Algorithm


INPUTS

L, , ,



set containing all nodes of network

M= Set containing all the nodes processed by the Algorithm



set of cognitive Radio nodes of network

set of primary radio nodes of network

state of primary radio user (OFF or ON)

= power of node v

= critical power (threshold)

= cost from source s to node v

= cost of link from v to w

=cost of link from v to x

// FIRST STAGE : Calculate the cost of every link.

for =1 to N

for =1 to N

for =1 to M


=.)
=

end for


end for

end for
// SECOND STAGE : Adjustment of the costs of links between primary radio sensor nodes and cognitive radio sensor nodes


for all nodes and in network N

if ( =ON)

for all neighbors of

C= 50// set cost of link to a very large value

else

C= original link cost from v to x


// THIRD STAGE: Load sharing technique

if ( Pv < Pc)

for all neighbors w of v

C = 50// set cost of link to a very large value

else


= original link cost from v to w

M = {s}

For all nodes v in network N except s

If v is adjacent to s



=

else


= ∞

While (M != N)

find a minimum D(W) for all w in (N – M)

For each v in (N –M) and adjacent to w, Calculate



Transmission Behavior of proposed efficient load sharing routing algorithm for CRSN


Practically, CRSN may consist of many small nodes but for our discussion and simulation we consider a multi-hop CR sensor network with only one primary user node and twelve CR nodes arbitrarily located on a plane. The CR users are able to operate on the licensed bands only if the primary user is not present or it is in the “OFF” state. When the primary user gets “ON”, all the CR nodes have to vacate the licensed bands as soon as possible. Node S is considered as the source node, node P as the primary user and rest of the nodes are used as the destination nodes. First stage of algorithm will calculate the cost of every link that contributes in determining which route should be preferred to get hold of the destination.

Lifetime of a network without load sharing algorithm


The nodes A, B and P are directly connected to source node S. We consider that source S transmit a large number of packets to random destinations. From figure 3 we can see that the cost of Link S-A is very low as compared to S-B and S-P i.e. C(S,A)<

Figure 3. A CRSN with 12 nodes including single primary user in “ON” state, Transmission is carried through node A.

As all the packets are transmitted through node A, the power of this node A (the power of the battery used in the node) decreases swiftly until it becomes zero and the node A becomes dead. Other nodes which are connected through this node A also get isolated. This causes transmission to stop. We can see the system’s output graphically, as shown in figure 4. A transmission is continued as long as node A is active.

Figure 4. Lifetime of a network without efficient load sharing technique


Load sharing algorithm to increase lifetime of a network when Primary user is “ON”


Life time of the above network can be increases by using proposed efficient load sharing technique. First we consider a network in which the primary user is in “ON” state and its band cannot be used that is represented by dotted line in Figure 3. When the primary user is in “ON” state the algorithm increase the cost of all connected links to very high value which is 50 in our case so that the transmission of packets cannot be happen through node P to avoid any interference.

From figure 5, It is clear that each node receives packet through the node A because of low cost path which causes the reduction in the power of A as shown in fig 2. As the power of node A becomes less than the threshold power, algorithm increases the cost of all the connected links to a very large value as shown in figure 4. Now the node A is used in power saving mode. If a link state algorithm is applied on the modified network, most of the destinations will be provided packet through node B as cost of A(S,B) is now minimum. Node A will be used only when there is no other path available as in case of node E, F and C. This will save a lot of power of node A and will increase the lifetime of overall network.

In power saving mode node A is only used to transmit packets to destinations E, F and C because these nodes can receive packets only through this node while the remaining nodes of system receive data packets through node B which has the lowest cost now. Figure 5 shows that in power saving mode, the cost of all the connected links is increased to a very large value 50. Figure 6 clearly demonstrate that at start the node A due to its large packet transmission rate behaves in active mode. But as the power of node A reaches threshold power it goes in the power saving mode and packet transmission rate decreases.


Figure 5. CRSN with node A in power saving mode. Transmission is carried through node B


Figure 6. Transmission behavior at node A with efficient load sharing technique

After node A enters power saving mode, node B will transmit packets to all the sensors except E,F and C as shown in the figure 6. As most of the packets are now routed through node B the power of node B decreases continuously.


Load sharing algorithm to increase lifetime of a network when Primary user is “OFF”

Consider the case when primary user is in “OFF” state and its band is used for transmission of data packets by secondary users. As the primary user is off, the cost of all the links connected to node P is decreased. As discussed earlier that when power of node A is reduced to a certain threshold, the data is transmitted through node B. When the power of node B reaches its threshold value, it is also declared as a critical node and turns into power saving mode. The cost of each link associated with the node B i.e. (A,B), (S,B) and (P,B) is raised to 50. Hence the load of all the other nodes except C, F, E, G and H is routed through node P. Now node B will send packets only to nodes G and H. Power saving mode as shown in figure 7.



Figure 7. Node B enters power saving mode and set all the connected links to a large value, Transmission is carried through node P.

Figure 8 demonstrate the transmission behavior at node B. At start, most of the packets are routed through node A therefore at this time the transmission rate of node B is low. As node A enters the power saving mode most of the packets are routed through node B which increases the transmission rate of node B. As the power of node B reaches threshold value, it enters power saving mode and is used only when there is no substitute path available for G and H therefore transmission rate is once again decreased.


Figure 8. Transmission behavior at node B with efficient load sharing technique

When node A and node B both enters the Power saving mode, the whole transmission is carried through node P. Figure 8(a) demonstrate the transmission behavior of primary node P whereas the figure 8(b) demonstrate the combined transmission of node A, B and P. At start, most of the packets are routed through node A therefore at this time the transmission rate of node B is low. As node A enters the power saving mode most of the packets are routed through node B which increases the transmission rate of node B. As the power of node B reaches power saving mode most of the packets are routed through node P which increases the transmission rate of node P and finally the node P reaches its power saving mode and is used only when there is no substitute path available therefore transmission rate is again decreased.


Figure 8(a). Transmission behavior at node A with efficient load sharing technique

Figure 8(b). Combined transmission behavior of node A, B and P with efficient load sharing technique

Similarly the same scheme is applied for every node. There is a clear increase in the lifetime of a network with every node implementing load sharing algorithm. With source A transmitting a large number of packets to random destinations, a transmission behavior of node I,E, F and G is shown in figure 9. The graph clearly demonstrates the load sharing phenomenon that occur among these nodes.


Figure 9. Transmission behavior at node I,E, F and G with efficient load sharing technique.


Conclusion


In this paper we have presented an efficient load sharing algorithm for cognitive radio sensor networks. The proposed algorithm defines a novel cost of link according to the characteristics and then apply the load sharing technique. A comparison is performed between original and proposed routing algorithm with the help of computer simulations. Our result clearly indicates that modified link state algorithm increases the lifetime of a network. Besides choosing the shortest path, the proposed technique also increases lifetime of a network by defining power saving mode. A small increase in transmission time is the only drawback of efficient load sharing routing algorithm.
References

R. Rubenstein, “Radios get smart”, IEEE Spectrum, Feb. 2007.

J. Mitola, “Software radios: Survey, critical evaluation and future directions,” IEEE Aerospace and Electronic Systems Magazine, vol. 8, pp. 25–36, Apr. 1993.

Federal Communications Commision, ”Spectrum policy task force report,” Report, Et docket No. 02-135, Nov. 2002.

J. Yang, Spatial channel characterization for cognitive radios, MS Thesis, UC Berkeley, 2004.

I. F. Akyildiz, W. Lee, M. Vuran, and M. Shantidev, ”NeXt generation/ dynamic spectrum access/ cognitive radio wireless networks: a survey”, Computer Networks Journal (Elsevier), vol. 50, pp.2127-2159, Sep. 2006.

SAFECOM, http://www.safecomprogram.gov/SAFECOM/.

Joint Tactical Radio System, http://enterprise.spawar.navy.mil/.

K. Zeng, W. Lou, and H. Zhai, ”On End-to-end Throughput of Opportunistic Routing in Multirate and Multihop Wireless Networks,” In Proc. of INFOCOM 08’, pp. 816-824, Apr. 2008.

D. Couto, D. Aguayo, J. Bicket, and R. Morris, ”A highthroughput path metric for multi-hop wireless routing,” In Proc. of ACM/IEEE MobiCom, Sep. 2003.

J. Zhao, R. Govindan, ”Understanding packet delivery performance in dense wireless sensor networks,” In Proc. of ACM Sensys03, Los Angeles, Nov. 2003.

H. Khalife, S. Ahuja, N. Malouch, and M. Krunz. “Probabilistic Path Selection in Opportunistic Cognitive Radio Networks” in Proceedings of the IEEE GLOBECOM, 2008.

Want R, Farkas K.I, Narayanaswami C, “Energy Harvesting and Conservation,” IEEE Pervasive Computing, , vol. 4, issue I, January-March 2005.

J. Choi., B. Y. Choi, Sejun Song, Kwang-Hui Lee, "NQAR: Network Quality Aware Routing in Error-Prone Wireless Sensor Networks," EURASIP Journal on Wireless Communications and Networking, vol. 2010. April. 2010





Download 38.98 Kb.

Share with your friends:




The database is protected by copyright ©ininet.org 2024
send message

    Main page