Joint Resource Allocation and Sensing Scheduling for

1 Australasian Telecommunication Networks and Applications Conference (ATNAC) 2010 - Mobile & Wireless Networks

Joint Resource Allocation and Sensing Scheduling for Cognitive Ultra Wideband 1,2

Norazizah Mohd Aripin, 1Rozeha A. Rashid and 1Norsheila Fisal, 1L.A.Latiff, 1S. H. Syed Ariffin, 1S.K. SyedYusof, 3Anthony Lo, 1M. Adib Sarijari 1

Faculty of Electrical Engineering Universiti Teknologi Malaysia, 81310 Skudai, Johor Bharu, Malaysia 2 Department of Electronics & Communication Engineering, Univeriti Tenaga Nasional, 43009 Kajang, Selangor 3 Wireless & Mobile Communications Group, Delft University of Technology, Netherlands [email protected], {rozeha| sheila| sharifah| kamilah| adib_sairi}@fke.utm.my, [email protected], [email protected] Abstract— Cross layer design is considered as an attractive strategy to enable multimedia transmission over cognitive ultra wide band (C-UWB) system experiencing time-varying channel conditions and traffic characteristics. The close dependency among various parameters across protocol stacks need to be addressed to determine the optimum cross layer approach. Hence, this paper highlights a MAC centric cross layer approach that is aware of sensing activity and channel conditions at the PHY layer, resource allocation at the MAC layer and MPEG-4 video at the APP layer. The cross layer actions are determined through optimal sensing scheduling and dynamic packet reception rate (PRR) based resource allocation to improve video transmission quality. Two important sensing approaches, local and cooperative OR-rule are also discussed in detail. Then, the impact of the proposed sensing scheduling and resource allocation to the video transmission are evaluated accordingly. Index Terms—Resource allocation, spectrum sensing, sensing time, cognitive radio, cross layer design

design is proven to perform well in wired networks. However, doing so is not suitable in wireless networks due to the dynamics in wireless medium. Main objective of CLD approaches are to enhance the performance of a system by jointly designing multiple protocol layers [2]. However, careful strategies and procedure are required to avoid undesirable consequences [3]. Hence, formal procedure to establish optimal initialization, grouping of strategies at different stages (i.e., which strategies should be optimized jointly), and ordering (i.e., which strategies should be optimized first) should be chosen appropriately. To overcome those challenges, numerous cross layer optimization methods [5-7] were proposed to provide opportunities for significant performance improvement by selectively exploiting the interactions between layers. Cross layer optimization methods can be categorized into application adaptation, applicationcentric adaptation, middle layer centric approach, middlewarebased adaptation and autonomous adaptation [4].

I. INTRODUCTION

W

IRELESS services have moved well beyond the classical voice centric cellular system and demand for wireless multimedia application is continuously increasing. Access to radio spectrum is regulated either as licensed or unlicensed. In licensed spectrum, the right for exclusive use is granted to individual operator. While in unlicensed spectrum, certain bands are declared open and can be accessed by any individual or operator following specific rule. Due to impressive success of unlicensed services and inefficient utilization of licensed spectrum, cognitive radio (CR) technology is now becoming an emerging technology ventured by industry players and research community. Each CR device must have the capability to sense their environment, learn about their radio resources and user/application requirements, and adapt behavior by optimizing their own performance in response to user requests [1]. The key challenge in enabling multimedia transmission over cognitive ultra wide band (C-UWB) networks lies on the dynamic allocation of network resources among users experiencing time-varying channel conditions and heterogenous video traffic characteristics. Traditional layered This project is funded by Research Management Centre (RMC), Universiti Teknologi Malaysia and Ministry of Science, Technology & Innovation (MOSTI) Malaysia.

Fig 1. Example of GOP structure with predictive dependencies

Predictive encoding technique such as MPEG-4, H.263 and H.264/AVC exploits spatial redundancy and temporal redundancy by using intra-frame and inter-frame coding respectively. A coded video sequence consists of a sequence of Group Of Picture (GOP) as shown in Figure 1 as one example. Each GOP consists of three types of video frames namely (i) Intra-frame (I-frame) (ii) P-frame that is dependent to the previous I-frame and (iii) B-frame that is bi-directionally dependent to the previous and following I or P frame. Considering MPEG-4 video transmission at the APP layer, the impact of losing I-frame on the received video quality is more significant than P or B frames due to video frame dependencies. While at the PHY layer, each users experience time varying channel condition that could also impact the

2

overall performance of multimedia transmission. This nature of dynamism and dependencies can be utilized by the MAC layer to further improve the performance. For example, MAC should schedule the video packet optimally based on its priority, dependency and delay deadlines. Additionally, each user is assigned with dynamic time slot allocation in accordance to their instantaneous channel condition and target packet reception rate. While at the APP layer, rate adaptation can be deployed to guarantee the QoS can be maintained within certain acceptable limit. Although substantial research has been carried out in the PHY layer perspective of CR system, this paper attempts to extend the existing research paradigm up to MAC and APP layer, which can be considered as pre-mature at this time. Quite recently, several research efforts are carried out to investigate the impact of sensing mechanism to the upper layer performance as in [8-11]. The proposed cross layer strategy consists of adaptive rate adaptation, optimal time slot allocation and sensing time scheduling particularly to improve video transmission quality. This paper is structured as follows. Section II describes our proposed UWB cross layer framework as well as the system model. In Section III, we highlight issues on local sensing, cooperative sensing and determination of optimal sensing time. Then, PRR based resource allocation is explained in Section IV. Section V presents our results and analysis, followed with conclusion and future recommendation in Section VI.

Based on the instantaneous channel condition, central controller calculates the optimal resource allocation (time slots) in accordance to the target packet reception rate (PRR) as well as target bit error rate (BER). The target PRR and BER are set to 8% and 10-6 respectively to meet the QoS requirement for multimedia application.

Fig. 2: Proposed cross layer interactions Central Controller (RX)

Node (TX)

Start

Set Attributes of Video Frame

Local Sensing

II. COGNITIVE UWB CROSS LAYER FRAMEWORK

Decision on Cooperative Sensing

Y

The proposed cross layer framework followed our previous works in [12,13] with time-division multiple access (TDMA) and only single hop transmission. Using TDMA, each user is assigned an optimal time slots for data transmission and channel sensing in accordance to their channel conditions. The current experienced dynamics at each layer is described by the state as shown in Figure 2. APP layer forwards its current state information on video frame priority, delay deadlines and dependency pattern to the MAC layer. At the same time, MAC also receives the current channel condition (represented by SNR) and also the appropriate sensing time. Based on the received information and its own queue status, MAC will determine the optimal time slot allocation, quantization level and schedule the packet accordingly. The decisions are then forwarded to the respective layers for actions. The interactions between C-UWB transmitter and the central controller (common receiver) are depicted as in Figure 3. The figure also illustrates message exchange activity between the two entities in general. At the start of every GOP, C-UWB users trigger the central controller about its intention to perform local sensing. With the objective of achieving high opportunistic access to C-UWB and high protection to PU users, cooperative OR-rule is investigated in this paper. The CUWB users report their sensing information to the central controller to be fused for the final decision of PU presence.

New GOP

N

Optimize Q-level

Compute Resource Allocation (EESM/PRR)

Update Resource Allocation

Optimize Scheduling Policy Update Time Slot Allocation Transmit Data

N End of Time slot?

Y New GOP? Y N End

Fig. 3: Interactions between C-UWB and central controller

Thus, the central controller can directly announce the allocated time slots and optimal data rate without have to wait for channel time request from C-UWB. Then, packet transmission is performed based on optimal scheduling policy

3

that resides at the MAC layer of C-UWB device. For simplicity, we adopt a round robin scheduling policy in this paper. Although it is a round robin mechanism, the MAC scheduling is improved by assigning different time slot allocation to each C-UWB nodes depending on their target and instantaneous PRR and BER, queue status and channel conditions of all C-UWB nodes in the network. Each C-UWB user is also assigned with an optimal sensing time to meet the target probability of detection (Pd) and probability of false alarm (Pf) during worst case channel conditions. Thus, the MAC scheduling is considered optimal in terms of the sufficient sensing time and the time slot allocation. III. LOCAL SENSING VS. COOPERATIVE SENSING For future opportunistic spectrum sharing scenarios, spectrum sensing is believed to be the key enabling technology to identify vacant channel for secondary spectrum access while limiting harmful interference to licensed primary users (PUs) [12]. Therefore, secondary or cognitive users (CUs) need to have reliable sensing mechanism to check the availability of a spectrum and change the radio parameters to exploit the unused spectrum The performance of spectrum sensing is characterized by the probability of detection, Pd which is the probability of correctly detecting the presence of a PU when it is active and the probability of false alarm, Pf, which is defined as the probability of wrongly detecting the presence of PUs while they are inactive [14]. Low Pf should be targeted to offer more chances of access for CUs. On the other hand, a high Pd will give a better protection for PUs [15]. In local sensing, each CU senses the spectrum within its geographical location and makes a decision on the presence of PUs based on its own local sensing measurements. Three wellknown local sensing techniques based on transmitter detection are matched filter detection, energy detection, and feature detection. Due to its coherent detection, matched filter is optimal detector if CU has a priori knowledge of PU signal. In the case of energy detection, the presence of PU is decided if the collected energy exceeds the pre-assumed threshold. This method is very simple since it assumes no information about the detected channel. The drawback is that it cannot reliably detect the signals with low Signal to Noise Ratio (SNR). Feature detection exploits the inherent periodicity in the received signal to detect primary signals with a particular modulation type [16]. It requires prior knowledge of the signals hence it can discriminate the noise well at the expense of long observation time and complex computation. In practical scenario, we cannot assume any predefined knowledge about the detected signal. Therefore, energy detector has been chosen as a candidate for this research. Transmitter detection is non cooperative detection, hence, it is vulnerable to hidden nodes problem as well as fading effects. It has been reported in [17] that cooperative detection can improve the performance of spectrum sensing by combining the observations of multiple CUs. It can reduce the sensitivity requirement of each CU in channel impairments such as shadowing, fading and low SNR conditions.

Cooperative sensing can also reduce detection time compared to individual sensing and thus, improves the agility of detection. The collected sensing results of each CU will be fused based on OR/AND-rule to reach a global decision on the occupancy of the channel in a cooperative network. This research is limited to hard fusion as it is shown in [18] that the performance of hard fusion schemes is comparable to that of soft fusion. Furthermore, the energy cost of sending one bit per decision is smaller than sending multiple bits per decision in soft decision scheme. The cooperation used can be deployed in a centralized or distributed manner. Work by Brodersen, et al. [18] has proposed a centralized scheme where an access point collects sensing results from all users. It then sounds the channel and performs channel allocation that meets the requested data rates of each user. In the distributed cooperation scheme as proposed by Cabric, et al. [19], neighbors are chosen randomly. Although the implementation maybe easier, it does not achieve the capacity that of centralized scheme. In this paper, a centralized cooperative sensing is chosen with fusion based on OR-rule as it improves the probability of detection and thus guarantees higher PU protection. A. Channel Sensing Hypotheses The sampled received signal, X[n] at the CU receiver will have two hypotheses as follows: 𝑊𝑊[𝑛𝑛] under 𝐻𝐻0 (PU is absent) (1) 𝑋𝑋[𝑛𝑛] = � 𝑊𝑊[𝑛𝑛] + 𝑆𝑆[𝑛𝑛] under 𝐻𝐻1 (PU is present) where n = 1, …, N; N is the number of samples. The noise W[n] is assumed to be additive white Gaussian (AWGN) with zero mean and variance σw2. S[n] is the primary user’s signal and is assumed to be a random Gaussian process with zero mean and variance σx2. Energy detector at each CU calculates the accumulated energy over T0 observation samples resulting in a decision statistic Y, which is compared with a threshold, γ, to decide whether signal is present or not. 2 (2) 𝑌𝑌 = ∑𝑁𝑁 𝑛𝑛=1(𝑋𝑋[𝑛𝑛]) Based on optimal decision yield by the likelihood ratio Neyman-Pearson hypothesis testing [20], Pd and Pf can now be defined as the probabilities that the CU’s sensing algorithm detects a PU under H0 and H1, respectively. Pf = P(Y > γ | H0) (3) Pd = P(Y > γ | H1) Since we are interested in low signal-to-noise ratio of PU (SNR= σx2/ σw2) regime, large number of samples should be used. Then

𝑃𝑃𝑓𝑓 = 𝑄𝑄 �

Thus, Pd is derived to be

𝑃𝑃𝑑𝑑 = 𝑄𝑄 �

2 𝛾𝛾−𝑁𝑁𝜎𝜎𝑤𝑤 4 �2𝑁𝑁𝜎𝜎𝑤𝑤

�

2 +𝜎𝜎 2 � 𝛾𝛾−𝑁𝑁�𝜎𝜎𝑤𝑤 𝑥𝑥

2 +𝜎𝜎 2 �2 �2𝑁𝑁�𝜎𝜎𝑤𝑤 𝑥𝑥

(4)

�

(5)

4

𝑄𝑄 −1 �𝑃𝑃𝑓𝑓 �−𝑆𝑆𝑆𝑆𝑆𝑆�𝑁𝑁�2

𝑃𝑃𝑑𝑑 = 𝑄𝑄 �

1+𝑆𝑆𝑆𝑆𝑆𝑆

�

(6)

While the number of samples, N, is the product of sensing time times sampling frequency and Q is the generalized Marcum Qfunction. B. OR-Rule Data Fusion In a cooperative spectrum sensing system using OR-Rule, the primary user is considered to be present if any of the cognitive radios detects the presence of the primary user. Assuming that there are M identical and independent cognitive radios in the cooperative spectrum sensing system, the cooperative probability of detection Qd and probability of false alarm Qf using OR-Rule are given by [24]: 𝑄𝑄𝑑𝑑 = 1 − ∏𝑀𝑀 (7) 𝑖𝑖=1�1 − 𝑃𝑃𝑑𝑑,𝑖𝑖 � 𝑄𝑄𝑓𝑓 = 1 − ∏𝑀𝑀 �1 − 𝑃𝑃 � (8) 𝑓𝑓,𝑖𝑖 𝑖𝑖=1 where Pd and Pf are respectively the probability of detection and probability of false alarm of a stand-alone cognitive radio. IV. PRR BASED RESOURCE ALLOCATIONS The PRR based resource allocation is deployed at the central controller through exponential effective signal to noise ratio mapping (EESM) technique. The basic idea of EESM is to find a compression function that maps a sequence of varying SINRs to a single value that is strongly correlated with the actual BER. Then, the estimated PER can be calculated directly as follows; (9) PER1(L)= 1-(1-BER)L Where L is the packet size and PRR is equal to (1-PER). In UWB multiband OFDM case, one channel is divided into three sub-bands (Ns = 1, 2, 3) and the allocation is made by subband; that means that each user is dynamically allocated one sub-band for the duration of one superframe. Therefore, the effective SINR calculated for each sub-band (SINRi) is given by [22]; 1

𝑁𝑁𝑠𝑠 𝑒𝑒 𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑒𝑒𝑒𝑒𝑒𝑒 = −𝜆𝜆𝜆𝜆𝜆𝜆 � ∑𝑖𝑖=1 𝑁𝑁

−𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆 𝜆𝜆

�

(10)

Knowing the target PRR (PRRT) and the instantaneous PRR (PRRi) of each user, central controller will compute the optimal time slot allocation. It is worth to note that the resource allocation is computed for each superframe. Let’s denote K as superframe size, the PRR based resource allocation can be described as follows; Input: PRRT1, PRRT2, PRR1, PRR2, K Output: Channel time allocation reserved for user 1 (N1) and user 2 (N2) Optimization Problem: How to increase the throughput of each user (meaning maximizing N1+N2) while maintaining a target PRR. Max (N1+N2) subject to: N1PRR1 ≥ PRRT1 N2PRR2 ≥ PRRT2 N1+N2 = K

The above optimization problem involves both inequality constraints as well as equality constraint. However, Lagrange optimization is only valid for equality constraint. Hence, the formulation has to meet the Karush- Kuhn Tucker (KKT) ‘complementary slackness’ condition to be simplified to Lagrange conditions. Complementary slackness implies that if the KKT multipliers are set to zero, then the corresponding inequality constraints must be negative for a minimum. From equation (11), we observed that when α, β (KKT multipliers) are set to zero, the inequality constraints become negative. Therefore, the optimization problem can be solved using Lagrange optimization as follows; L= (N1+N2)+α(PRRT1-N1PRR1) +β(PRRT2-N2PRR2) (11) +γ(N1+N2-K) with γ being Lagrange multiplier. Derivatives of L set to zero yield; (12) δL/δN1 =1- α PRR1 (13) δL/δN2 =1- β PRR2 (14) δL/δ α = PRRT1 – N1PRR1 (15) δL/δ β = PRRT2 – N2PRR2 (16) δL/δ γ = N1+N2 – K Let’s define the ratio of PRR targets of both users as; (17) a= PRRT2/PRRT1 Thus; (18) N1=K*(1-PRR2)/(a*(1-PRR1)+(1-PRR2)) (19) N2=K*a*(1-PRR1)/(a*(1-PRR1)+(1-PRR2)) Using the same approach, the resource allocation can be extended to M-multi user case. For variable number of user M, we can approximately estimate each user i is allocated with; 𝐾𝐾 𝑀𝑀𝑖𝑖 = 𝑗𝑗 =𝑁𝑁 (1 − 𝑃𝑃𝑃𝑃𝑃𝑃𝑖𝑖 ) 1 + ∑𝑗𝑗 =0 (1 − 𝑃𝑃𝑃𝑃𝑃𝑃𝑗𝑗 ,𝑗𝑗 #𝑖𝑖 ) (20) In the next section, further insights of this algorithm using simulations will be explained. V. SIMULATION RESULTS & ANALYSIS In this section, we present the simulation results of our proposed MAC-centric cross layer design. The simulation is performed using MATLAB and Network Simulator 2 (NS-2). The impact on the received video at the APP layer is also presented. Figure 4 to 6 demonstrates the performance of OR-rule cooperative sensing in terms of probability of detection and probability of false alarm. It can be observed from Figure 4 that PU detection is greatly enhanced by OR-rule cooperative sensing as it improves the probability of detection under various SNR conditions. However, probability of false alarms also increases as shown in Figure 5 and hence, reduces the opportunistic access for CUs. Furthermore, under a bad SNR condition (-7 dB), there is basically no access allowed for CUs as the probability of false alarms approaches almost 100%. Therefore, it is recommended that opportunistic access for CUs is allowed only in good SNR condition. Since the probability of false alarms recorded by individual node is much lower than by cooperative sensing, it is also recommended that attempts of

5

spectrum access is carried out based on local sensing rather than cooperation. Cooperative Sensing_OR rule 1

Qd

0.8 0.6 0.4

higher throughput can be achieved. While during bad channel condition, lower data rate is assigned to minimize packet error rate to maintain the target BER. Using QPSK modulation as standardized in [23], the threshold limit for SINR is estimated through simulation and summarized as in Table 1. This threshold is then will be used in our adaptive rate allocation. SINR for BER of 10-4 is also presented because some multimedia applications can tolerate up to this bit error.

0.2

Cooperative Sensing-OR rule (Pd=90%, SNR=-6 dB)

0 1

2

3

4

SNR=-3 dB

5

user

0.8

SNR=-7 dB Qf

SNR=0 dB

1

Fig. 4: Performance of cooperative sensing using OR-rule

0.6 0.4 0.2 0

Qf

Cooperative Sensing-OR rule (Pd=90%, n=10)

1

1 0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0

2

n=200

3

4 n=400

5

user

Fig. 6: Performance of probability of false alarm using OR-rule cooperative sensing under various SNR conditions

1

2 SNR=0 dB

3 SNR=-3 dB

4

5

user

SNR=-6 dB

Fig. 5: Performance of probability of false alarm using OR-rule cooperative sensing under various SNR conditions

Figure 6 illustrates that the disadvantages of OR-rule cooperative sensing in terms of probability of false alarms can be significantly overcome by using more samples for detection and hence, longer sensing time. In the case of the set Pd is 90%, for all 5 users and bad SNR condition of -6 dB, a target Qf of about 10% can be achieved by using sample size of 400. To evaluate the impact of sensing activity (and thus additional delay) to multimedia application, simulations were carried out using Foreman video input and each user is allocated 14µsec to perform local energy sensing. Job failure rate (JFR) was used as performance metric to represent how many packets were lost as compared to the whole packets generated. Packet loss may be due to delay deadline or corrupted during transmission. Figure 7 depicts that the video quality degrades when sensing activity is included at the MAC superframe. Interestingly, the quality degradation is quite minimal because the ratio of sensing period to superframe size is small. In short, C-UWB users can gain benefit of awareness to channel condition to perform adaptive resource allocation with only insignificant overhead. Hence, only minimal delay is introduced due to overhead and this is preferred especially in multimedia applications. To ensure a high quality received video, the target BER is set to 10-6. However, this QoS requirement is difficult to maintain especially under dynamic wireless channel conditions. To overcome this, user shall utilize good channel conditions by allowing higher transmission rate and thus

Fig. 7: Video quality performance with and without sensing

Data Rate (Mbps) 53.3 80 110

Table 1: SNR Threshold Limit Threshold SINR Threshold SINR @ BER=10-4 @ BER=10-6 -3dB -6dB -1.9dB -4.3dB -0.5dB -2.9dB

Modulation (code rate) QPSK (1/3) QPSK (1/2) QPSK(1/3)

Figure 8 shows the timeslot allocation assigned by PNC to two users based on their target PRR as well as instantaneous PRR. To clearly appreciate the performance, one of the users is set fixed (good/bad) and the other one is set random. As depicted in Figure 8, more time slots are allocated to the user with good channel condition. In contrast, user with a very bad channel condition is not given any timeslot access. Using this technique for video transmission, the impact to video quality is quite different than in data transmission. User with bad channel condition may experiences throughput reduction. However, as long as the I-frame (the most important frame) is granted with resource and transmission, the impact to perceived video quality is still acceptable. Figure 9 shows the video performance in terms of average JFR when our cross layer design approach that is aware of the dynamic channel condition and APP layer QoS target was implemented. We compare our cross layer design approach

6

with the non-cross layer design (non-CLD In the non-CLD case, each user is assigned fix time slot all the time regardless of channel condition. Note that our proposed cross layer design outperformed the non-CLD. However, the non-CLD performs better when more than 8 users share the limited resources. This is due to the fact that when more users are competing, the user that experience bad channel condition will always get the minimum timeslot as compared to the fix timeslot allocation.

REFERENCES [1]

[2] [3] [4]

[5]

[6] [7]

[8] Fig. 8(a)

[9]

[10]

[11]

[12] Fig.8(b) Fig. 8(a),(b): PRR based resource allocation

[13]

Job Failure Rate (%)

35 30

[14]

25 20 15

[15]

10 5

[16]

0 2

4

non-CLD

6

proposed CLD

8

10 Users

Fig. 9: Impact of proposed cross layer design to received video quality

[17]

[18]

VI. CONCLUSIONS In conclusion, SNR is the main QoS metric at the PHY layer to determine the appropriate sensing time for cognitive users. Cooperative sensing enhances decision making by collaborating CUs. To allow higher protection to PU, OR-rule cooperation is recommended. Optimal time slot for optimal resource allocation must be assigned for sensing activity and data transmission at the MAC layer. Simulation results had shown that with appropriate sensing time allocation, the target probability of detection can be achieved. Beside, the impact to video transmission is not significant due to optimal sensing time allocation, minimal overhead and also delay.

[19]

[20]

[21] [22]

[23]

A. Batra, S. Lingam, J. Balakrishnan, “Multi-band OFDM: a cognitive radio for UWB ”, in IEEE International Symposium on Circuits and Systems, 2006 (ISCAS 2006),pp. 4094-4097, 2006. E. Setton et. al, “Cross Layer Design of Adhoc Networks for Real Time Video Streaming”, IEEE Wireless Communications, August 2005 V. Kawadia & P.R Kumar, “A Cautionary Perspective of Cross-Layer Design”, in IEEE Wireless Communications, February 2005. F. Fu & M. van der Schaar, “A New Systematic Framework for Autonomous Cross-layer Optimization”, IEEE Transactions on Vehicular Technology, Vol. 58, No. 4, May 2009 N. El-Fishawy, M.Shokair & W. Saad, ‘Proposed MAC Protocol Versus IEEE802.15.3a For Multimedia Transmission Over UWB Networks’, Progress in Electromagnetics Research B, Vol 2, 189-206, 2008 H. Gu and C. Yang, “Power and Admission Control for UWB Cognitive Radio Networks”, IEEE Communication Magazine , 2008 H. Shiang & M. Schaar, ‘Queing-Based Dynamic Channel Selection for Heterogeneous Multimedia Applications Over Cognitive Radio Networks’, IEEE Transactions on Multimedia, 2008 A. T Hong & Y. Liang, “Adaptive Scheduling of Spectrum Sensing Periods in Cognitive Radio Networks”, IEEE GLOBECOM 2007 H. Kushwara et. al, “Reliable Multimedia Transmission Over Cognitive Radio Networks Using Fountain Codes”, Proceedings of the IEEE, Vol 96. No 1 January 2008 M. P Wylie-Green, “Dynamic Spectrum Sensing by Multiband OFDM radio for Interference Mitigation”, IEEE International Symposium on New Frontiers in Dynamic Spectrum Access Networks (DySPAN), 2005 N. M. Aripin et.al, “Evaluation of Required Sensing Time for Multimedia Transmission Over Cognitive Ultra Wideband System”, IEEE International Conference on Ultra Modern Technology (ICUMT), October 2009. Rozeha A. Rashid, et al., “Cross Layer Design of Cognitive Radio MBOFDM System”, 3rd International Conference on Postgraduate Education (ICPE3), Penang Dec 12 – 14 2008 Norazizah Mohd Aripin, et. al, “Cross Layer Design of Multimedia Transmission Over Cognitive Radio UWB Multiband OFDM System’, Int’l Graduate Conference in Engineering & Science (IGCES), Johor Bharu, Dec 2008 M. Höyhtyä, A. Hekkala, and A. Mämmelä, ”Spectrum Awareness: Techniques and Challenges for Active Spectrum Sensing” in Towards Cognitive and Cooperative Wireless Networking: Techniques, Methodologies and Prospects, pp. 353-372, Springer, 2007. A. Sahai, N. Hoven, R. Tandra, “Some Fundamental Limits on Cognitive Radio”, Proc. of Allerton Conference, Monticello, 2004Asff Čabrić et al., “Spectrum Sharing Radios,” IEEE Circuits and Systems Magazine, vol.6, pp. 30–45, 2006. A. Ghasemi and E. S. Sousa, “Collaborative Spectrum Sensing for Opportunistic Spectrum Access in Fading Environment”, Proc. IEEE DySPAN, Nov 2005 R.W Brodersen, A. Wolisz, D. Cabric, S.M. Mishra, D. Wilkomm, 2004, White Paper: CORVUS – A Cognitive Approach for Usage of Virtual Unlicensed Spectrum, available online http://www.bwrc.eecs.berkeley.edu/MCMA Cabric, D, Mishra, S.M & Brodersen, R.W., “Implementation Issues in Spectrum Sensing for Cognitive Radios”, Asilomar Conference on Signals, Systems and Computers, 2004 A.A. El-Saleh, M. Ismail, O. B. A. Ghafoor, A. H. Ibrahim, “Comparison Between Overlay Cognitive Radio And Underlay Cognitive Ultra Wideband Radio For Wireless Communications”, Proc. Of 5th IASTED Int’l Conf. Communication and Networks (AsiaCSN2008), Langkawi, Malaysia, 2-4 April 2008 K.B. Letaief and W. Zhang, ”Cooperative Spectrum Sensing”, book chapter in Cognitive Wireless Communication Networks, Springer, 2007 A. Khalil M. Russiere & J-F. Herald, “Dynamic Cross-Layer Spectrum Allocation for Multi-Band High-Rate UWB Systems”, Lecture Notes in Electrical Engineering 41, Springer, 2009 Standard ECMA-368: High Rate Ultra Wideband PHY and MAC Standard, 2nd Edition, Dec 2007