Spectral Management for a Cognitive Radio Application with Adaptive Modulation and Coding

Transcription

1 445 Spectral Management for a Cognitive Radio Application with Adaptive Modulation and Coding Mohammed Amine Azza 1 *, Ali El Moussati 1, Slimane Mekaoui 2,Kamal Ghoumid 1 1 Signals, Systems and Information Processing team, National School of Applied Sciences, Oujda, Morocco 2 Telecommunications Department, Faculty of Electronics and Informatics, USTHB, Algiers, Algeria a.elmoussati@ump.ma, smekaoui@yahoo.fr, ghoumid_kamal@yahoo.fr Abstract Cognitive Radio allows Software Defined Radio terminal to perceive its environment and then interact with it. In other words, cognitive radio can collect information from its surroundings, model them and try to adapt its behavior according to them. The main aim of this paper is to shed light on the concept of spectrum sensing and adaptive cognitive radio and propose an application implemented on Small Form Factor (SFF) Software Defined Radio (SDR) development platform. This application includes both a spectrum sensing algorithm based on Fast Fourier Transform which is basically insensitive to noise level and an Adaptive Modulation and Coding (AMC) scheme. The validation of the proposed cognitive radio application and the efficiency of the AMC and the spectrum sensing algorithms are shown through experiments and measurement results. Index Terms- Spectrum Sensing, Cognitive radio, Software Defined Radio,Adaplive modulation and coding, SDR Platform. 1. I. INTRODUCTION Radios that sense all or part of their environment are considered aware systems. Awareness may drive only a simple protocol decision or may provide network information to maintain a radio s status as aware. A radio must additionally autonomously modify its operating parameters to be considered adaptive. When a radio is aware, adaptive, and learns, it is a Cognitive Radio CR [1]. Among the operating parameters that may be adapted on cognitive radio systems, we find: Frequency, instantaneous bandwidth, modulation scheme, error correction coding, channel mitigation strategies such as equalizers or RAKE filters, system timing (e.g., a Time Division Multiple Access (TDMA) structure), data rate (baud timing), transmit power, and even filtering characteristics. In the same way, Frequency adaptation (spectrum sensing), modulation and coding adaptation are referred to several research and various applications, where the spectrum sensing is the task upon which the entire operations of cognitive radio rests, even less, a few studies dealing adaptability in cognitive radio especially the adaptive modulation and coding, thus the combination of these two principles in one application remain the goal of our recent work. Our goal in this work is to develop in a single cognitive radio implementation a better spectrum management, and a better adaptation in terms of modulation and coding. To this end, we propose, in this paper an application involving spectrum management with a spectrum sensing algorithm and adaptive modulation and coding scheme for whatever be the communication channel chosen. The paper is organized as follows: The first section summarizes all the theory behind these two principles namely spectrum management, its technical and adaptive modulation and coding scheme, in the second section we present the experimental results and steps of implementation, in the last section, we discuss the experimental results.

2 446 II. THEORY AND PRINCIPLES Cognitive radio and adaptation is now the trend in all applications of wireless communications, this trend is based on the desire to obtain a communicating terminal that has a conscience towards his environment and an enormous capacity to respond and adapt. As we mentioned earlier, cognitive radio can be described as a node in a network that can sense his operating environment and adapt his implementation to achieve the best performance. The principle of this application, included in the IEEE and IEEE h standard, requires an alternative spectrum management which is as follows: the Secondary User (SU) may at any time access to frequency bands that are free, not occupied by the Primary User (PU) having a license on this band. The SU will assign it once the service ended or after a PU has shown these attempts to connection. This application allows also to the terminal at any time to have the capability of using multiple modulations and choosing the most appropriate modulation and coding depending on the linkk quality (IEEE a standard), this concept is known as Adaptive Modulation and Coding AMC [2]. In the following, we will try to present the whole theory behind this application, these two objectives, the algorithm used for spectrum sensing, and then we will detail the intelligence cycle ensuring adaptation in terms of modulation and coding. 1. A. Algorithm for spectrum sensing For the spectrum sensing, there are many possible techniques, such as energy detection, matched filter detection, cyclostationary feature detection, covariance-based detection, and wavelet-based detection [3][4]. In our case, we chose to use of the energy detection which guarantees a relatively increased level of efficiency, the output of which gives the decision variable. This variable is then compared with a threshold and if it is above the threshold, then the result of the detector is that a primary user is present. Energy detection is very useful since it requires no detected needed signal. For this purpose, we have adopted a blind spectrum sensing algorithm, called fast Fourier transform (FFT)-averaging-ratio (FAR), and proposed as depicted in Figure 1 given below. Segmentation & Windowing Fig.1. Flow chart of FAR algorithm The input to the FAR algorithm [5][6] is a baseband discrete-timee signal sampled at frequency (fs), while the output is a series of vectors of two-class decisions that represent the availabilities of the channel. The input signal is in real numbers. Firstly, in each time slot, a block of base-band signal samples are segmented into T frames. Denote t-th frame of the input samples by X t (n), the segmented frames are multiplied by a window function: X w,t (n) = X t (n).w(n) (1) N : is the number of samples in a frame. T : is the number of frames. Then, the FFT is applied to the windowed frame. FFT The power spectral density (PSD) calculation follows the FFT operation. P X & PSD Computation The PSDs of T consecutive frames are used for averaging, yielding: information about the Ratio computation! " (2) 0$$N&1 (3) (4) & Thresholding

3 447 Let P m be the mean of P avg (k) calculated across all frequency tones. ) * (5) In order to be robust to the noise level, the decision variable r(k) is formed as a ratio. + / - (6) Finally, thresholding is applied to r(k) for k = 0, 1,..., N/2, and the decisions on channel states are made according to the following rule: +. / : / ;2832<3 (7) B. Adaptive modulation and coding. The main concept of adaptive coding and modulation is to maintain a constant performance by varying transmitted power level, modulation scheme, coding rate or any combination of these schemes.this allows us to vary the data rate without sacrificing BER performance. thresholds are carefully selected; the system can then reach the highest performance. [7] (Fig.2). We used the physical layer of the a standard [8] as adaptive implementation of our system on the SFF SDR platform. The Table 1 shows for each mode the modulation type, rate coding, throughput, the relative error of constellation and Error Vector Magnitude. Table 1: modulation and coding dependent parameters in IEEE a standard Mode Modulation Code Rate Data rate (Mbps) Relative constellation error (db) EVM (% rms) 1 BPSK 1/ BPSK 3/ QPSK 1/ QPSK 3/ QAM 1/ QAM 3/ QAM 2/ QAM 3/ The system consists of a transmitter, a receiver and a Rayleigh communication channel figure 3. Fig.2. Variation of SNR versus time and choice mode based on thresholds The selection of modulation mode for the next transmission heavily depends on the current channel quality estimation. If the channel quality can be measured accurately, ideal switching between different modes is available. Under these circumstances and if, the switching Fig.3. Block diagram of IEEE a transceiver architecture The transmission part consists in a block for generating the binary data and a block which contains multiple modulation and coding schemes to select the correct pattern. The OFDM time signal is generated by an inverse FFT and is transmitted over the Rayleigh fading channel after the cyclic extension has been inserted. The received signal is serial to parallel converted and passed to a FFT operator, which converts the

4 448 signal hack to the frequency domain. This frequency domain signal is coherently demodulated. Then the binary data is decoded by the Viterbi hard decoding algorithm. Signal to noise ratio is estimated at receiver and then transmitted to the transmitter through feedback channel. Then, the transmitter according to the estimated SNR selects the appropriate modulation scheme and the coding rate that maintains a constant bit error rate lower than the requested BER [9]. III. IMPLEMENTATION OF SPECTRUM SENSING AND (AMC) ON SFF SDR PLATEFORM The cognitive radio transceiver with a spectrum sensing algorithm and adaptive modulation and coding is implemented on the SFF SDR DP [10], which is composed of three functional modules: radio frequency (RF) module, data conversion module, and digital processing module, constitute the SFF SDR DP. The application development is made with the help of Simulink from Matlab Source Software, this type of development provides more robustness and relevance. A. Hardware interface. The platform is built around the digital processing module. The latter is designed around the TMS320DM6446 (also called DM6446) Digital Media Processor (DMP) System on Chip (SoC) [5] from TI and Virtex-IV XC4VX35 FPGA from Xilinx. DM6446 combines an Advanced Very Long Instruction Word (VLIW) 64x+ DSP and Reduced Instruction Set Computer (RISC) ARM926J-S cores, where the ARM microcontroller is mainly set to run the INTEGRITY Real-Time Operating System (RTOS) while DSP performs complex data processing. The data conversion module is equipped with a 125 MSPS, 14- bit dual channel ADC and a 500 MSPS 16-bit dual channel interpolating DAC provided by TI. The RF module is configured to have either 5 or 20 MHz bandwidth with working frequencies of MHz for the transmitter and MHz for the receiver. The cognitive radio application implementation (Figure 1) is divided into different tasks each consisting of several modules. The SFF SDR platform gives the designer the option to choose the silicon device that is most suitable to the task being developed. We use the INTEGRITY and SMSHELL API provided by Lyrtech to target the board while we develop our signal processing tasks on the DSP core and the FPGA. The division of tasks between the DSP and the FPGA was made based on the availability of resources, the inherent characteristics of these cores, and the extra functionalities offered by TI and Xilinx. We used the readily available Xilinx Logicore Blocksets for FPGA and the optimized DSP libraries written for vectors of complex numbers for C64x+ core. Interfacing between the DSP and FPGA is done using the Video Processing Sub-system (VPSS) data port. The VPSS consists of a Video Processing Front End (VPFE) and a Video Processing Back End (VPBE). B. Software interface of FAR algorithm. Included with the hardware is the board support design kit (BSDK) that includes the software drivers to support TI Code Composer Studio (CCStudio) Integrated Development Environment (IDE) and Green Hills INTEGRITY RTOS and MULT IDE for TMS320DM6446 DSP SoC and Xilinx ISE foundation development tool for the FPGA. Support for these tools through the BSDK enables application development partitioned and targeted independently for the DSP SoC and the FPGA. The support for high-level model based software design flow using the model-based design kit (MBDK) allows for ease of development and easy partitioning of, software functions across a multiprocessing architecture and takes the board support package to the next level. The SDR DP includes both the BSDK and the MBDK that enables seamless integration of The MathWorks model-based design tools to the lower level DSP tools (CCStudio) and FPGA tools (ISE Foundation). With the model-based flow,

5 449 INTERNATIONALIONAL JOURNAL OF MICROWAVE AND OPTICAL TECHNOLOGY, developers can use either C/HDL or MATLAB to rapidly develop and test proof-of-concept designs and then optimize the architecture for cost and power for a specific application. The distribution of the FAR algorithm and the adaptive modulation and coding cycle components between the DSP core and the FPGA is shown in Figure 4. The VPBE and VPFE are used to transfer the data streams back and forth between the two modules while the custom registers are used for handshaking [10]. to the DAC. The second subsystem is used to implement the bit based operations, such as convolution encoding and interleaving. Finally, the Interpolation block synchronizes the IFFT output with the DAC operating at 120 MHz. DAC_I Filter DAC_Q Interpolation Modulation Process_Control Fec_interleaver IFFT Fig.4. System data of the sensing application. C. FPGA design for Adaptive modulation and coding. We implemented the IEEE a transceiver on the Lyrtech Small Form Factor (SFF) Software Defined Radio (SDR) Development Platform. The receive path of the Lyrtech RF module down-converts the received signal to an intermediate frequency at 30 MHz. This means that the received IEEE802.11a signal at the input of ADC has a maximum frequency of 40 MHz and so the minimum sampling rate should be 80 MHz in accordance with the Shannon-Nyquist criteria. The system generator [11] implementation of the transmitter is divided into five subsystems as shown in the Fig. 6. The first subsystem which is named as process Control is designed for controlling the signal flow from the MAC layer Fig.5.. System Generator model of the transmitter The system generator implementation of the receiver is divided into nine subsystems as shown in Fig. 6. The first subsystem, named as Down_Converter, converts the IF signal received from the ADC into base-band I and Q components. Then, the frame detection, timing and frequency synchronization processes are performed by the preamble_decode subsystem. After obtaining the frequency domain subcarrier symbols, the chan_equalizer subsystem performs the channel equalization and carrier phase tracking algorithms. The Viterbi_Decoder subsystem decodes the bit stream that has been encoded by convolution codes.. Finally, the decoded bits, except for the signal field, are fed to the descrambler subsystem to reconstruct the transmitted message. The decoded bits of the Signal Field are also sent to the Adaptive_Control block to extract the frame parameters such as modulation type, number of transmitted bytes and the number of OFDM symbols in the frame.

6 450 Fig.6. System Generator model of the receiver IV. EXPERIMENT DEMONSTRATION AND RESULTS DISCUSSION The SFF SDR platform is considered as a secondary user. The frequency band covered by the RF module of the platform is divided into a variety of channels, and a spectrum detection algorithm is adopted for detecting the availability of each channel. according to the detection results of the channel, a signal with a carrier frequency belongs to the unoccupied channel is sent simultaneously by considering the signal to noise ratio of the channel as a metric, for detecting the primary user input to the channel. If this is the case, the secondary user must assign the channel, and the carrier frequency of the emitted signal will be changed by another belonging to an available channel (frequency hopping is performed). The process repeated until the end of the transmission. In the demonstration, an FRS handset (used in the frequency band of 462 Mhz) and a talkie- band of walkie PMR446 (used in the frequency 446 Mhz) are used as PUs call each other, While the platform sends a signal said secondary with carrier frequency of 545 Mhz, and one SFF SDR DP running FAR algorithm senses the spectrum in real-time. The experiment is conducted in an indoor environment as illustrated in figure 7. As mentioned before, the platform is considered as being a secondary user transmission and a receiver at a time. Below, we review all possible cases for the three channels, and the results obtained in each case. Fig.7. Schematic illustrating the use scenario. The figure 8 is a compound figure that represents the power spectral density of the 3 signals belonging to the 3 channels (Channel 1: FRS 462 Mhz, Channel 2: PMR 446 Mhz, Channel 3: 545 Mhz). These three signals are derived from our implemented transmitterr and received by the receiving antenna via our designed receiver which acts in this case as a spectrum analyzer (The colors used in the figure 8 are intended to tell the difference between the signals). Fig.8. Compilation of the power spectral density of the three separately received signals. The use of the power spectral density (DSP) is necessary, because it is the unit used in calculating the ratio r(k) to apply thresholding as noted above in the definition of the FAR algorithm. Note clearly the presence of a peak in the transmission frequency band that has a relatively greater power spectral density, that makes the ratio r(k) greater than the detection threshold.

7 451 In the remainder of the testing phase, we have tried to transmit a signal in a channel (See the example of figure 9: Channel 1: FRS 462 Mhz) in the presence of the signal from the PMR (Channel 2: PMR 446 Mhz). Fig.10. BER Performance of each mode in Dispersive channel with practical estimation. Fig.9. The power spectral density of the received signals. The result of the spectrum sensing in this case indicates the occupation of these two channels, then, the FRS is used to inject a signal which will be received by the platform, this causes tilting of the carrier frequency to another channel judged free channel. After examination of all possible cases in the proposed scenario, we note that the algorithm has all the expected results, because the objective was to exploit opportunistically the spectrum holes that are detected automatically in conjunction with the transmission. In order to test our IEEE802.11a physical layer prototype on the Lyrtech SFF SDR Development Platform, we also developed a simple MAC layer running on the platform and a user interface program on the PC. We measured the BER performance of our prototype just for 5 modes (1, 3, 5, 7, 8) as in these cases we have the better performance of Bit Error Rate. The obtained BER curves are depicted in Fig. 10 for these modulation types with the associated coding rates. We can see in Figure 10 that below 11 db, no mode gives a lower BER Table 2: MCS depends of SNR interval thresholds SNR values (db) [-,7] [7,18] [18,29] [29,34] [34,+ ] MCS Mode 1 Mode 3 Mode 5 Mode 7 Mode 8 In Figure 11 measurements of the BER as a function of signal to noise ratio at the reception are illustrated. In the same Figure we have the throughput depending on the SNR. Fig.11. BER and throughput performance of Adaptive Coded Modulation in Dispersive channel with practical estimation

8 452 From figure 11 we can observe that in the case of adaptive modulation and with an SNR level greater than 11 db is selected, our experiment allows keeping a BER lower than 10-3 by choosing the best and most appropriate coefficient for the spectral efficiency. These results were already confirmed by simulations done in previous studies [12]. Consider a channel that has a deep high level of fading, the options here are to use one of five modulation modes, which differ in spectral efficiency and robustness. If the fading is considered to be extremely deep, perhaps half of all bits will be interpreted as error bits. Here, it is advantageous to send fewer bits because the total number of errors will be decreased, which influences bit error rates much more than total number of bits sent. When the channel is not in the case of a fading, then many bits are wanted to be sent. In this situation, the BER is lowered by increasing the number of bits sent because errors become less frequent. It is the combination of these two principles that allows the BER performance of adaptive systems to be more robust than static systems while simultaneously providing better spectral efficiency at most ranges of SNR. V. CONCLUSION In this paper, the implementation of an application that combines two main objectives of the cognitive radio spectrum sensing and adaptive modulation and coding is presented. The selection of solutions for critical processing task is made; the FPGA implementation of the IEEE802.11a physical layer on the Xilinx Virtex- 4 sx35 chip is detailed. We have analyzed the performance of energy detection based spectrum sensing techniques using FAR algorithm, compared to other energy detection techniques,the FAR algorithm has clearly better performance in identifying spectral holes between spectrally well-contained PUs. Moreover, a real-time demonstration of spectrum sensing and adaptive modulation and coding has been conducted, and very encouraging experiment results have been obtained. However, main project goals were achieved and solid grounds for the future research. REFERENCES [1] J. Polson, Cognitive Radio Applications in Software Defined Radio, in SDR Forum Technical Conference and Product Exposition, November 15 18, [2] Nevio Benvenuto and Filippo Tosato, On the selection Of Adaptive Modulation and Coding Modes Over OFDM, IEEE Communications Society, [3] S. Haykin, D. Thomson, and J. Reed, Spectrum sensing for cognitive radio, Proceedings of the IEEE, vol. 97, no. 5, pp , May [4] T. Yucek and H. Arslan, A survey of spectrum sensing algorithms for cognitive radio applications, IEEE Communications Surveys & Tutorials, March [5] M. A. Azza, A. El Moussati and R. Barrak, Implementation of Cognitive Radio Applications on a Software Defined Radio Plateform in ICMCS'14. Marrakech. April [6] Zhe Chen, Nan Guo, and Robert C.Qiu, Demonstration of real-time spectrum sensing for cognitive radio IEEE Communications Letters, vol. 14,no. 10,pp ,2010. [7] A. Homayoun and B. Razavi, "A 5-GHz 11.6-mW CMOS receiver for IEEE a applications" in CICC IEEE, [8] IEEE b, Part 11: Wireless LAN Medium Access Control (MAC) and Physical Layer (PHY) Specifications, Sept [9] Hyung Suk Chu, Byung Su Park and Chong Koo An,"Wireless Image Transmission based on Adaptive OFDM System",IEEE, [10] Lyrtech SFF SDR development platform technical specs, Lyrtech Inc. [11] Xilinx Inc., "XtremeDSP for Virtex-4 FPGAs UserGuide." entation/user_guides/ug073.pdf, [12] S. N. Abdullah, Z. M. Abid, Adaptive Coded Modulation for OFDM System, Journal of Engineering, Vol. 2, Feb