Energy Detection Technique in Cognitive Radio System

Transcription

1 International Journal of Engineering & Technology IJET-IJENS Vol:13 No:05 69 Energy Detection Technique in Cognitive Radio System M.H Mohamad Faculty of Electronic and Computer Engineering Universiti Teknikal Malaysia Melaka Melaka,Malaysia Abstract Energy detection is one of the spectrum sensing techniques in cognitive radio system that facilitate to detect vacancies in spectrum frequency. Energy detection also denoted as non-coherent detection is the signal detection mechanism employing a power detector to determine the availability or nonavailability of signal in the band. In this paper, BPSK modulation signal is generated randomly as the input signal to the energy detector to analyze the spectrum frequency. Probability of detection is obtained throughout the detection process using various numbers of samples and SNR value. Index Term- component; Cognitive radio; energy detector detection; BPSK signal. I. INTRODUCTION Currently, wireless communication system faced a big challenged as there are many electronic devices use wireless as the transmission medium. The growth of this system involved more spectrum resources since there are a lot of devices use multimedia applications for communication. The rapid development of this technology will cause congestion in allocated spectrum band and leading to the user dissatisfaction. In the meantime the use of each spectrum frequency is dissimilar in term of the number of user and the usage time. These will create opportunities in spectrum frequency that is allocating to licensed user. Moreover there are a lot of licensed bands have been found to be grossly underutilized such as television broadcasting, amateur radio and paging [1]. These vacancies open to a latest alternative in spectrum utilization that is cognitive radio (CR). Cognitive radio technology is a radio transmitter that can alter the parameters based on interaction with the environment in which it operates []. The main objective of cognitive radio is to get the best frequency which is not used based on its ability to detect the environment. It can improve the utilization of spectrum by allowing secondary user to make use of unused spectrum from primary user or to share the spectrum with the primary users. Frequency utilization in wireless communication systems that use cognitive radio is categorized into: i. Primary user (PU) - User that have high priority in the use of frequencies. ii. Norairin Mahmat Sani Faculty of Electronic and Computer Engineering Universiti Teknikal Malaysia Melaka Melaka,Malaysia Secondary User (SU) users are given less priority in the use of frequencies as compared to primary users. Secondary user that used the frequency spectrum must not cause interference to the primary user and as well not cause disruption to the aircraft, satellites, local and goods carrier that is operated accordance to the ITU Radio Regulations. Secondary user also can claim protection from interference that cause by other secondary user that using the same frequency band. The purposes of this research are to determine the primary user through probability of detection using different number of samples and SNR values. The rest of the paper is organized as follows. In section II, the concept of cognitive radio is discussed. A description of the system model and the simulation is shown in Section III. In Section IV are the results of the spectrum utilization in the specified area and the analysis. Finally in Section V is the conclusion of this paper. II. A. Spectrum sensing CONCEPT OF COGNITIVE RADIO Spectrum sensing is the most important components in cognitive radio as its ability to sense and aware the parameters related to the radio channel characteristics. Spectrum sensing acts to detect the presence or absence of a primary user signal in cognitive radio system [3]. This element enables SU to access unoccupied spectrum band. The fundamental nature of spectrum sensing is a binary hypothesis-testing problem [4]: Ho: primary user is absent H 1 : primary user is in operation Meanwhile the input metric in the spectrum sensing is given by [5]: i. Probability of correct detection, P d which quantifies the probability of a SU detecting that incumbent is present ii. Probability of false alarm, P f - which quantifies the probability of SU declaring that a incumbent is

2 iii. International Journal of Engineering & Technology IJET-IJENS Vol:13 No:05 70 present in the spectrum when the spectrum is in fact x(t) Band Pass Squaring Integrator Threshold free. filter device Probability of miss detection, P m - which quantifies the probability of SU declaring that the spectrum is free but the fact is there is incumbent present. Fig. 1. Block diagram of energy detection The probability of correct detection, P d and probability of false alarm, P f is important for the evaluation of detection performance is defined as (1) [6]: P d P decision H1 H1 P H1 (1) P f P decision H1 H o P H o Based on Figure 1 the input signal x(t) is filtered with a band pass filter in order to limit the noise and to select the bandwidth of interest. The noise in the output of the filter has a band-limited, flat spectral density. The power, from integrator is then compare with threshold value to examine the two hypotheses H o and H 1 H o or H 1 Where is the decision statistic and is the decision threshold. Technique of spectrum sensing can be categorized into three groups which are [7]: i. Transmitter detection the detection based a signal from primary transmitter through the local interpretation of SUs. ii. Cooperative detection SUs shares their sensing information and combined the decision for a better and precise detection. iii. Interference-based detection detection using interference temperature model. B. Transmitter Detection Techniques Transmitter detection is based on the discovery of a weak signal from a primary transmitter through the local observations of CR users. The basic hypothesis model for transmitter detection technique is defined as () [7]: w( n)... H o x( n) y( n) w( n)... H1 Where x(n) is the received signal, y(n) is the transmitted signal and w(n) is the noise. H o is a null hypothesis states that there is no PU in a certain frequency band. While H 1 indicates that PU is exists in that frequency band. There are three different techniques that generally used in transmitter detection which are matched filter detection, energy detector and cyclostationary feature detection [6]. One of the simplest techniques that decides the present and absent of PU based on the energy of the observed signal is energy detector [1]. Not as matched filter, this technique doesn t need any priori information from PU. () Matched filter is designed to maximize the output signal to noise ratio for a given input signal in digital signal processing presence of additive stochastic noise. This filter correlates the signal with time shifted version and compares the final output with the predetermined threshold [8]. Based on the block diagram show in Figure, the primary user signal x(t) is convolved with a time shift version h(t) of the prior known signal s(t). The final output of the filter is compared with the threshold will determine the primary user presence. x(t) AWGN Channel h ( t) s( T t ) Mixed Signal Matched Filter Fig.. Block diagram of matched filter detection Cyclostationary is the most complex technique in spectrum sensing among the three mentioned above. This detection technique makes use of the periodicity of received PU signal to spot the presence of PUs signal [6]. This technique can distinguish the PU from interference and noise. III. SSTEM MODEL Generally, to obtain the probability of detection, P d in specific channel will involve two main processes which are BPSK signal and energy detector stage. The input signal is generated randomly using BPSK modulated signal. The input signal is generated based on the number of samples, N. While the energy detector stage engaged 4 main blocks to produce the detection of PU. A. Binary Phase Shift Keying (BPSK) signal Threshold BPSK modulation is a two state period shift keying whereas the phase of the message signal is set to 0 or according to the value of the modulating signal. The modulating signal is a binary sequence and multiplied alongside a sinusoidal carrier, thereby the BPSK modulated signal was generated [9]. If a 1 is sent, the modulated signal looks precisely as the carrier alongside 0º initial phase. If 0 is

3 International Journal of Engineering & Technology IJET-IJENS Vol:13 No:05 71 sent, the modulated signal has erupted of carrier alongside 180 initial phase. The transmitted signal for BPSK modulation is shown in Figure 3 [9]. 1 N 1 T ( y[ n]) (6) N n 0 Fig. 3. Principle of BPSK modulation and demodulation The transmitted signal for BPSK modulation is: s1( t) Acos( fct ) S i ( t) s( t) Acos( fct ) Where s 1 (t) and s (t) represent two electrical signals, A is amplitude of modulating signal and f c is the carrier frequency. After executing linear combinations of two orthonormal basis function between ϕ 1 (t) and ϕ (t), only one is needed [10]: (3) s ( t) 1( t) cos( fct) ( t) (4) EBPSK Tb Every bit of the modulating signal causes a transmitting symbol with T s duration identical with the bit duration T b. Each signal from (3) is of duration T b seconds and has finite energy given below [9]: Tb EBPSK A (5) B. Energy Detector Stage Where T is the decision variable, y[n] is the received signal, N is the number of samples. The integrator computes the energy, of the detected signal across a specific observation interval. Finally, this output signal is compared to the threshold in order to decide whether a signal is present or not. The threshold is set according to statistical properties of the output when only noise is present. The equation of probability of detection is given as (7) [10]: Pd Pr { i H1} Qu(, i ) (7) Where γ represent SNR, and λ i is the threshold for i th signal sample. The fundamental substitution between probability of miss detection, P m =1-P d while P f has different implication in the context of dynamic spectrum sharing. IV. RESULT AND ANALSIS The result display in this part will consist of two parts which are the generation of BPSK signal and detection of spectrum holes from energy detection technique. Each output from each energy detection block will be analyzed in detail in this section. A. Generation of BPSK modulated signal The generation of BPSK signal is run in different number of samples which are N=8 and N=16. Figure 5 below show the original digital signal was from the random binary data input 0 or 1. The binary data input was computed by the BPSK equation which produced a BPSK signal with two phase shifts. Then the BPSK signal multiplied with the carrier signal which produced the result signal. To compute the signal power in particular frequency band in time domain, a band pass filter (BPF) is requested to the target signal and the power of the signal samples is measured. Figure 4 depicts the implementation of energy detector. To evaluate the power of the received signal, the output of BPF is squared and integrated over an interval T. s(t) Band Pass filter (.) ( ) dt 0 Threshold Fig. 4. Implementation of energy detector The decision static for energy detector is:

4 International Journal of Engineering & Technology IJET-IJENS Vol:13 No:05 7 Fig. 6. Output from band pass filter The output from the band pass filter was directed to the squaring device. In this stage, the peak and time of the output filter waveform were increased and become positive value. The output of the squaring device was shown in Figure 7. Fig. 8. Complementary ROC under AWGN for (a) SNR =-10 db; (b) SNR-0 db Fig. 7. Output from squaring device Performance of energy detector can be described through complementary receiver operating characteristic (ROC) curves. In [11] a ROC curve is used to compare the performances for different threshold values. The curve is between P d an P f show the relationship between sensitivity and specificity of sensing method for various threshold. In this paper, ROC curve is between P m vs P d as indicates in Figure 8. A high P m would result in missing the presence of PU, which then increases interference to PU. Figure 8 illustrates the output of P d and P m with different value of SNR. When the SNR value declined from -10 db to -0 db, P d value also decreased. The result in Table I show that that energy detection capable to detect spectrum from small to larger number of samples. When the value of SNR decrease P d will also decrease, means the system detect less number of channels. Table I Analysis performance of P d in various SNR and number of samples N SNR P d

5 International Journal of Engineering & Technology IJET-IJENS Vol:13 No:05 73 V. CONCLUSIONS In this paper energy detection method is presents with the BPSK modulated signal. Energy detection attained to sense the present channel with random BPSK signal in various SNR value and number of samples. Throughout the simulation, probability of detection, P d and probability of miss detection, P m were obtained and were presented in complementary ROC graph. As the number of SNR value decrease, the number of detection channel as well will lessen. The simulation also able to detect PU in various numbers of samples, N but the probability of detection, P d in a large number of N will slightly low than the smaller N. This paper will be enhances by choosing other modulation signal such as QPSK to be compared their performance in P d and P m. VI. ACKNOWLEDGEMENT The first author wishes to acknowledge the Universiti Teknikal Malaysia Melaka for giving her financial support through short term grant, PJP/011/FKEKK (36B)/S REFERENCES [1] Ma, J., Li, G.., & Juang, B. H. (009). Signal processing in cognitive radio. Proceedings of the IEEE, 97(5), [] Mitola, I. and Maguire, J. G.Q Cognitive Radio: Making Software Radios More Personal. IEEE Personal Commun. Mag. Vol 6, No. 4 :13-8. [3] Bhargavi, D., & Murthy, C. R. (010, June). Performance comparison of energy, matched-filter and cyclostationarity-based spectrum sensing. In Signal Processing Advances in Wireless Communications (SPAWC), 010 IEEE Eleventh International Workshop on (pp. 1-5). IEEE. [4] Zhang, W., Mallik, R. K., & Ben Letaief, K. (008, May). Cooperative spectrum sensing optimization in cognitive radio networks. In Communications, 008. ICC'08. IEEE International Conference on (pp ). IEEE. [5] Pratas, N., Prasad, N. R., Rodrigues, A., & Prasad, R. (011, February). Cooperative spectrum sensing: State of the art review. In Wireless Communication, Vehicular Technology, Information Theory and Aerospace & Electronic Systems Technology (Wireless VITAE), 011 nd International Conference on (pp. 1-6). IEEE. [6] Akyildiz, I. F., Lo, B. F., & Balakrishnan, R. (011). Cooperative spectrum sensing in cognitive radio networks: A survey. Physical Communication, 4(1), [7] Kapoor, S., & Singh, G. (011, February). Non-cooperative spectrum sensing: A hybrid model approach. In Devices and Communications (ICDeCom), 011 International Conference on (pp. 1-5). IEEE. [8] Ziafat, S., Ejaz, W., & Jamal, H. (011, August). Spectrum sensing techniques for cognitive radio networks: Performance analysis. In Intelligent Radio for Future Personal Terminals (IMWS-IRFPT), 011 IEEE MTT-S International Microwave Workshop Series on (pp. 1-4). IEEE. [9] Popescu, S. O., & Gontean, A. S. (011, October). Performance comparison of the BPSK and QPSK Modulation Techniques on FPGA. In Design and Technology in Electronic Packaging (SIITME), 011 IEEE 17th International Symposium for (pp ). IEEE. [10] Digham, F. F., Alouini, M. S., & Simon, M. K. (007). On the energy detection of unknown signals over fading channels. Communications, IEEE Transactions on, 55(1), 1-4. [11] ucek, T., & Arslan, H. (009). A survey of spectrum sensing algorithms for cognitive radio applications. Communications Surveys & Tutorials, IEEE, 11(1),