Reduce The Handoff Latency In Cognitive Radio Network Computer Science Essay

Published: Last Edited:

This essay has been submitted by a student. This is not an example of the work written by our professional essay writers.

Due to the enormous development in communication technology, demand for the spectrum is expected to grow even more tremendously in the coming years. To cope up with the ongoing demand, one of the most promising concepts to facilitate the flexible usage of the radio environment is knows as Cognitive Radio. Disaster scenario is the most amazing usage case of cognitive radio networks, where CR node can use the best available spectrum to establish an ad-hoc network through its cognition capability and reconfigurability. However, with most of the spectrum being already allocated, it is becoming exceedingly hard to find unoccupied bands either to deploy new services or to enhance the existing one due to the lack of frequency bands in radio spectrum. In this case, the most important challenge is to share the licensed spectrum without interfering with transmission of other licensed users. Hence, CR node should immediately vacate the spectrum upon detection of licensed user on the particular frequency band, which known as spectrum handoff or spectrum mobility. In order to manage the spectrum handoff, current researches are mainly focus to stay in licensed spectrum. In addition, there are some other proposals that take in consideration of both licensed and unlicensed frequency band. However, most of the cases they manages the channel state information in static manner which it is very much impractical due to dynamic nature of the spectrum allocation. The idea of my work is to manage the information about the availability of spectrum in such a way that CR users can reduce the spectrum handoff. To achieve that, my focus is to develop an algorithm to discover the spectrum opportunities as fast as possible to incorporate with DSA. The benefits and drawbacks of such strategies will be compared with more conventional approaches.  A mixture of simulation and analysis will be used to assess performance of the algorithm. Game theory and Markov analysis will be particularly important analytical tools for the spectrum selection process among the secondary users.

Table of Contents 3

1 Introduction 4

2 State of Art / Related Work 4

3 Open Research Question 5

4 Solution Ideas 6

5 Schedule 6

6 Context to the Graduate School 7

7 Interface with Other Research Topics within Graduate School 8

8 Demonstrator 8

9 References 9


The increasing demand for new wireless services and applications, as well as the increasing demand for higher capacity wireless networks, the wireless networks become highly heterogeneous, with mobile devices consisting of multiple radio interfaces. In this context, it is essential to have updated information on radio environment to enhance the overall network performance. The outcomes of several investigations have shown that the lack of spectrum is not an issue, but the fact that radio resources are used inefficiently. Therefore, a promising functionality is required to be built into future terminals to have the cognitive capability to assist with the Dynamic Spectrum Allocation (DSA), which allows more efficient utilization of radio resources by changing the spectrum allocation on demand. A cognitive radio is a self-aware communication system that efficiently uses spectrum in an intelligent way [1]. The most significant characteristic of a cognitive radio is the capability to sense surrounding radio environment and make a decision to adapt the parameters for maintaining the quality of service. It autonomously coordinates the usage of spectrum in identifying unused radio spectrum on the basis of observing spectrum usage. Therefore, spectrum handoff occurs when a licensed user further utilizes this unused radio spectrum and find that CR nodes occupy the channel [2]. Beside this, there are other two main factors, which may trigger spectrum handoff: 1) lost of connectivity due to the mobility of nodes involved in an on-going communication 2) degradation of QoS in current spectrum band. The operational procedure of Spectrum Handoff involves two main phases: primary user detection and link maintenance. Once there is a primary user (PU)/licensed user in the specific portion of the spectrum used by secondary user (SU) or CR node, the CR node has to immediately release the spectrum for the PU. In order to avoid service termination, the CR will perform link maintenance procedure to reconstruct the communication. In general, link maintenance procedure can be categorized into a) proactive spectrum handoff b) reactive spectrum handoff. In proactive spectrum handoff, CR nodes observe all channels to obtain the channel usage statistics, and determine the candidate set of target channels for spectrum handoff while maintaining the current transmission and perform spectrum handoff before the link failure happens [3]. Reactive spectrum handoff operates in on-demand manner, i.e. CR nodes perform spectrum switching after detecting the link failure [4]. From system design, point of view reactive spectrum handoff is more suitable than the proactive spectrum handoff as it requires very complex algorithm for concurrent operation. On the other hand, proactive spectrum handoff offers very faster spectrum switching with respect to reactive spectrum handoff, resulting better QoS in on-going transmission.

State of Art / Related Work

Although the concept of cognitive radio technology sounds very simple, the design and implementation have large number issues to solve. The problem of spectrum handoff in cognitive radio network has been widely investigated in the last few years. L. Giupponi, in [5] proposed a fuzzy-based spectrum handoff to deal with the incompleteness, uncertainty and heterogeneity of a cognitive radio scenario. In [6], Chang and Liu proposed a strategy that optimally determines which channel to probe and when to transmit in a single channel transmission. A sensing sequence has proposed in [7] to maximize the chances of finding and idle channel but it does not guarantee the minimum discovery delay. Moreover in [8] authors proposed a Bayesian learning method to predict the unutilized radio spectrum. To enhance the proactive spectrum handoff, a concept of backup channel and candidate channel are introduced in IEEE 802.22 [9], but it is not shown how to construct backup and candidate channel list. In order to minimize the discovery delay a fast opportunity discovery mechanism has proposed by Kim and Shin in [10] where they build a sensing sequence that helps to find the available spectrum with minimum delay. In addition, an optimal and efficient algorithm is designed to update the backup channel list by importing and exporting channels from candidate channel. But most of the algorithm mainly concern to handoff in licensed band although CR nodes have the capability to use any portion of the spectrum not only from licensed band but also unlicensed band. Therefore authors in [11] first come up with an idea to build the backup channel list from unlicensed band in static manner and proposed a Markov channel model to evaluate the scheme. It is observed that, in these studies, the dropping probability and the number of handoff is reduced in case of the appearance of primary users. In order to reduce the spectrum handoff probability a spectrum matching algorithm for CR nodes is illustrated in [12]. To evaluate the spectrum handoff procedure four performance metrics are proposed in [13] and generalized the teletraffic parameters for both Pus and SUs to formulate the theoretical analysis. In the study [14] a non-cooperative game theoretic framework is proposed to evaluate the spectrum selection process among secondary users who can opportunistically select best spectrum opportunity, under tight constraint not to hamper primary licensed users. Cooperative game model is designed in [15] to analyze a multi-hop wireless scenario where nodes need to agree on a fair allocation of spectrum. In [16], the authors propose a resource allocation algorithm to maximize the spectrum efficiency, fairness and optimality by using the concept of Nash Bargaining solution.

Open Research Question

The main goal of the cognitive radio network is enhance the dynamic spectrum access by reuse the legacy spectrum dynamically and opportunistically without interrupting the primary user transmission. Although this concept enhances the utilization of radio spectrum but in the meanwhile it poses a lot of research challenge in network performance. One of the major and critical research challenges is spectrum handoff. Currently there exist a very few research effort to address the problem of spectrum handoff in cognitive radio network. In real time communication, CR node may experience a number of spectrum handoff to successfully complete the service. During each spectrum handoff, if there is an available channel for the CR node, the CR node can switch the channel without any delay. Otherwise the CR node has to wait for the period to discover the new channel. This latency is very significant in multimedia communication which is not only related to RF front-end but also on spectrum sensing, spectrum decision, link layer, and routing. So it is necessary to design an effective algorithm in cross layer approach in minimize the handoff latency. As we mentioned earlier of this proposal, spectrum handoff can be initiated for three different reasons, so it is essential to develop a flexible spectrum handoff framework to distinguish the different handoff strategies. Such as, proactive spectrum handoff can be use in case of delay sensitive application but for wireless sensor node which are more energy constrained will adopt reactive spectrum handoff. Therefore depending upon the application scenario the spectrum handoff strategies can be different. Link maintenance probability is another crucial factor which needs to investigate in very precise way. It may happen that CR node can co-exist with the PUs by reducing it transmission power level or adopting any efficient mitigation techniques, resulting to reduce the number of spectrum handoff.

Solution Ideas

The aim of this work is to enhance the spectrum handoff by reducing the number of handoff, minimize the handoff latency and in the same time develop a flexible spectrum handoff framework. Therefore my idea is to develop an algorithm which will dynamically create channel list by utilizing both licensed band and unlicensed band while CR nodes are in transmission, which will increase the probability to reduce the number of handoff. Figure 1 illustrated the concept of both vertical and horizontal handoff. In this scenario a dynamic non-cooperative game theory model will use to

Licensed Spectrum


Vertical Handoff

Horizontal Handoff

Unlicensed Spectrum



Fig.1: Vertical and Horizontal Handoff

discover the best available spectrum, which will further minimize the handoff latency. In the game model secondary users are the players and their strategy space is composed of available unutilized spectrum. As handoff latency is not only depends on spectrum discovery but also link layer and routing. In that case idea is to work on cross layer approach to enhance the overall handoff procedure. Beside this, several co-existence technique can be investigate to avoid the unnecessary spectrum handoff.


The total research schedule is subdivided in seven different phases and all the phases should be done one after another since there is a chronological relationship between the phases. A detailed work plan with the methods and estimated time is given in the table in the following section.


Nature of the task

Estimated Time (Month)

Literature Review, State of the Art Analysis

Problem Formulation


Proposed an algorithm to reduce the spectrum handoff latency, and number of spectrum handoff

Analyzed the proposed algorithm through Game theory and Markov chain Model


Evaluate the proposed algorithm through simulation in NSMIRACLE


Proposed a flexible spectrum framework in cross layer approach


Combined the algorithm proposed in phase 3 with the flexible spectrum framework


Evaluate the overall system performance through simulation


Finalized the work and paper works on Doctoral Thesis


Context to the Graduate School

The main research focus of the graduate school is to develop a self-configure mobile communication in disaster scenario taking the consideration of radio resources, protocol and network operation. In disaster scenario, the communication suddenly drops out encompassing a vast geographical area probably including some major communication nodes. Moreover, it may cause power failure results the devices connecting the terminals and network to cut off, and may end up with drastic capacity drops and serious communication blackouts. With the grace of cognition capabilities, the Cognitive radio can resolve these problems in disaster situations. For instance, various nodes can work as a relay to establish communication when there is a power outage or a loss of central control. Therefore, it is necessary to response the overall network as fast as possible. Spectrum handoff is one of the major issues that are account for delay. To make the system more reliable and efficient for disaster scenario it is obvious to minimize the number of handoff and handover latency.

Interface with Other Research Topics within Graduate School

There are several issues are investigated by different researcher within the graduate school in order to design a self-organized mobile communication in disaster scenario. Therefore it is necessary to have a strong collaboration between them to build the system as a whole. The aim of my project is to develop an algorithm in cross layer approach which will able to accelerate spectrum handoff.

Physical Layer Interface: Spectrum handoff is closely coupled with the physical layer information. In this case, the information about the available free channel through spectrum sensing is depends on physical layer activates. Moreover, the switching time to reconfigure the RF front-end is a major issue to minimize the handoff latency is depends on physical layer. In this aspect, the work of Rami Nicolas will helps us to enhance the overall spectrum handoff performance. The work of Jin Yang on radio resource, channel history and communication resources in certain geographical location could helpful to properly characterize the teletraffic pattern of primary and secondary users. One of our colleague from graduate school Noman Murtuza, he is working on the reconfigurable antenna design could also support to adopt a mitigation technique like detect and avoid method, Transmission power control, low duty cycle to avoid unnecessary spectrum handoff.

MAC Layer Interface: In CR network the most key function is to sense the radio environment to explore the spectrum holes or white space and to avoid interference with Pus. In order to share and access the channels sensed to be idle, the CR node must implement MAC protocol. MAC is responsible to investigate the order in which the spectrum bangs must be searched to minimize the time for finding the best available spectrum. Therefore the work of Saleh Hussin on Cognitive Radio MAC is tightly related with the spectrum handoff performance.

Routing Layer Interface: Apart from the presence of primary users, the spectrum handoff can also occur due to the node mobility and significant degradation of QoS on current communication link. In all cases after choosing a free channel it is necessary to discover the new routing path. To find a new routing path CR node need to perform neighbour discovery. The new routing approach by Thomas Finke could help to switch between multiple routing protocols to adjust routing in a very flexible way. 


To evaluate the performance of proposed cross layer architecture and algorithm will examined through simulation in NSMIRACLE environment. The earlier version of NS2 (network simulator 2) does not support multiple radio interfaces and lacks flexible tools for the cross-layer control of communication systems. In addition, in the standard distribution of the simulator, the wireless channel is represented via unrealistic models, which may lead to biased results. Therefore, starting from some primitive NS classes, some new classes that are the basic blocks to model heterogeneous architectures are defined. This allows to partially reusing the code developed for NS-2 with slight modifications. Moreover, it introduces standardization in the cross layer messaging in order to give to all the modules/algorithms the possibility of communicating to each others by means of ad-hoc structured messages.