CHANNEL MODELLING & PERFORMANCE ANALYSIS OF WIFI

CHANNEL MODELLING & PERFORMANCE ANALYSIS OF WIFI Jivisha, Gaurav Mohta, Saumya Das Siim Manipa Institute of Technoogy Emai Id- gauravmohta.officia@gmai.com; Ph No. +9 9609850304 Abstract Wi-Fi or wireess fideity is a wireess way to hande networing. It is aso nown as 80. networing. In this paper we have proposed the technique of channe modeing using a modified Saeh-Vaenzuea mode, using Naagami-m distribution of sma scae fading. The modeing is done for the foowing frequency ranges: for UWB channes covering frequency range from to 0 GHz. This range covers indoor residentia, indoor office, industria, outdoor and open outdoor environments with a contrast between LOS and NLOS properties. Through this paper 00 impuse responses are reaized for each environment. Keywords channe modeing, Compementary Code eying(cck), Saeh-Vaenzuea, Naagami-m distribution, Utra Wide Band (UWB) channes, Orthogona Frequency Division Mutipexing (OFDM) channes, Line of Sight (LOS), Non-Line of Site (NLOS) INTRODUCTION Wi-Fi or wireess fideity is a wireess way to hande networing. It is aso nown as 80. networing. Using this technoogy we can connect computers anywhere in a home or office without using wires. An average speed of 54 Mbps is provided. In 999 IEEE 80.b was introduced. 80.b defines the physica ayer and media access contro (MAC) sub ayer for communication across a shared WLAN. At physica ayer, 80.b operates at.45 GHz with maximum bit rate of Mbps. At MAC sub ayer, 80.b uses carrier sense mutipe access with coision avoidance (CSMA/CA) protoco. For IEEE 80.g standard uses.4ghz and provides a data rate of 54 Mbps. It specifies OFDM and CCK moduation schemes with 4Mbps as maximum mandatory data rate[3]. The disadvantages observed in 80.b and 80.g are that they use.4 GHz spectrum which is crowded with other devices such as Buetooth, microwave ovens, cordess phones, or video sender devices, among many others. This may cause degradation in performance. Aso Power consumption is fairy high compared to other standards, maing battery ife and heat a concern. With the aim of providing Very High Throughput (VHT) IEEE 80. ad was introduced in the year 003. Wi-Fi ad or WiGig defines a new physica ayer operating at 60 GHz wave spectrum [3]. The bandwidth aocation is between.4 GHz 5 GHz. 60 GHz wave transmissions wi scae the speed of WLANs and WPANs to 6.75 Gbits/s over a distance of 0 meters. It supports FST (Fast Session Transfer) protoco which maes it bacward compatibe with 5 GHz or.4 GHz[4]. For maing 80. ad compatibe at the Medium Access Contro (MAC) or Data Lin Layer, devices consists of three radios:.4 GHz for genera use which may suffer from interferences, 5 GHz for more robust and higher speed appications, and 60 GHz for utra-high-speed within a room. It aso supports session switching between the.4ghz, 5GHz and 60 GHZ unicensed band. The MAC protoco is based on TDMA and Physica ayer uses SC (Singe Carrier) and OFDM (Orthogona Frequency Division Mutipexing) to simutaneousy enabe ow power, high-performance appications [3]. In previous researches, the concept of arge scae fading, sma scae fading (Rayeigh fading) has been proposed. Large scae fading between the transmitter and receiver is predominanty affected by arge his, forests, buidings etc. Sma scae fading refers to change in signa ampitude and phase that is experienced as a resut of sma changes in between the transmitter and receiver. Sma scae fading is dependent on two factors, namey, time spreading of signa (signa dispersion) and time variant behavior of channe. For mobie appications, the channe is time variant because of the motion between transmitter and receiver resuts in change in propagation paths. The rate of change of these propagation conditions is caed fading rapidity. Sma-scae fading is aso caed Rayeigh fading because if the mutipe refective paths are arge in number and there is no ine-of-sight (LOS) signa component, the enveope of the received signa is statisticay described by a Rayeigh PDF. There are three basic factors that affect signa propagation in systems. They are refection, diffraction and scattering. []. Refection occurs when a propagating eectromagnetic wave impinges on a smooth surface with very arge dimensions compared to the RF signa waveength.. Diffraction occurs when the radio path between the transmitter and receiver is obstructed by a dense body with arge dimensions compared to. Diffraction is a phenomenon that accounts for RF energy traveing from transmitter to receiver without a ine of-sight path between the two. It is often termed shadowing because the diffracted fied can reach the receiver even when shadowed by an impenetrabe obstruction. 3. Scattering occurs when a radio wave impinges on either a arge rough surface or any surface whose dimensions are in the order of or ess, causing the refected energy to spread out (scatter) in a directions. Utra Wide Band (UWB) channes tends to push the imit of bandwidth to about 500 MHz or more or uses a bandwidth that is 0% or arger than the carrier frequency. We use UWB channe in this paper because of the property of UWB channes i.e. impuse responses of UWB can be sparse i.e. characterized by a few spies separated by time during which no significant energy arrives. The 695 www.ijergs.org

fundamenta mechanism of wireess communication is mutipath propagation. The eectromagnetic fieds sent from transmitted antenna are components of magnetic fieds transmitted in different directions []. Each of the components now propagates in space and might be refected, diffracted, or scattered by objects (mountains, houses, trees, was, furniture) in the environment, (see Fig.) Fig. MULTI PATH TRANSMISSION Each interaction process can change the direction of the components, and some interactions (ie diffraction) might even spit up the components into mutipe new components. This causes mutipath components. For this paper, we have designed a modified Saeh-Vaenzuea mode, consisting of Naagami-m distribution of sma scae fading, with different m-factors for different components [3]. MODELLING According to the recommendations of channe modeing subgroup of IEEE 80.5.4a, the tas group has to mandate an aternative physica ayer for sensor networs and simiar devices woring with the IEEE 80.5.4a MAC ayer. The main goa of the proposa was to deveop an energy efficient data communication with data rates between Kbps and severa MegaKBps and with the capabiity of geo-ocations. With the channe modeing, we try to mae a fair comparison of different proposas. The main goas of the mode described in the document are the modeing attenuation and deay dispersion. Modeing of attenuation incudes shadowing and average path oss and in deay dispersion power deay profie and sma scae fading statistics are incuded. From these other parameters such as rms deay spread, number of mutipath components carrying x% of the energy etc. are being considered [3]. The generic mode described for different frequency ranges are amost simiar but the parameterizations are different. A the modes discussed are time continuous and the impuse responses are being generated with the hep of MATLAB program and its being tested for different underying environments. We derived number of environments discussed for the evauation of the mode. The ist discussed is not comprehensive and cannot cover a possibe future appications []: i. Indoor residentia: Environment for home networing. Buidings consist of sma units with was of reasonabe thicness. ii. Indoor office: Due to partitioning of unit, higher attenuation is given by furniture, cubices, abs etc. iii. Industria environments: These are characterized by arger encosures (factory has), fied with a arge number of metaic refectors.this is anticipated to ead to severe mutipath. iv. Body-area networ (BAN): It is characterized by the wireess networ of wearabe computing devices. Channe mode for this is considered to be very different for this environment as the main scattering is in the near fied region of the antenna. v. Outdoor: Whie a arge number of different outdoor scenarios exist, the current mode covers ony a suburban-ie microce scenario, with a rather sma range. vi. Agricutura areas/farms: For those areas, few propagation obstaces (sios, anima pens), with arge distances in between, are present. Deay spread can thus be anticipated to be smaer than in other environments. The ey features for the generic channe mode are summarized beow before going into detais [] mode treats ony channe, whie antenna effects are to be modeed separatey d n aw for the pathoss frequency dependence of the pathoss modified Saeh-Vaenzuea mode: Arriva of paths in custers. Mixed Poisson distribution for ray arriva times. Possibe deay dependence of custer decay times. some NLOS environments first increase, then decrease of power deay profie referring Naagami-distribution of sma-scae fading, with different m-factors for different components Boc fading: Channe stays constant over data burst duration. 696 www.ijergs.org

SALEH VALENZUELA MODEL - It is described as a statistica mode where it is assumed that mutipath components (MPCs) arrive in custers which are formed by the mutipath refections from various components paced in the vicinity of transmitter and receiver. Different arriva rate of custers in the MPCs as we as inter custer arriva rates are described with the hep of Poisson Processes that are exponentiay distributed. The compex, impuse response described for this channe in genera is given by:- ( ) ( ) () th Where, is the arriva time of th custer and is the the arriva time of the th ray measured from the beginning of the custer, th th whie is the gain of the ray of the custer. The phases are uniformy distributed i.e., for a bandpass system, the phase is taen as a uniformy distributed random variabe from the range0., As we have earier mentioned that number of custers L in MPCs is assumed to be a Poisson distributed process which can be formuated by given equation:- ( ) ( ) ( ) () Where, L and competey characterizes the distribution. Within each custers, Custer arrivas are Poisson distributed with rate ʌ ( custer arriva rate) and ray arriva rates in the custer are aso Poisson distributes with rate λ (ray arriva rate) such that, λ >> ʌ. The definition assumes that within a custer, the first ray arrives with no deay ( ). The distribution of custer arriva time with the ray arriva time is described by given equations:- ( ) [ ( )] (3) ( ( ) ) [ ( ( ) )] (4) In this document to accord the variation in the fitting for the indoor residentia, and indoor and outdoor environments, we mode the ray arriva times as the mixtures of two Poisson processes as foows:- ( ( ) ) [ ( ( ) )] ( ) [ ( ( ) )] (5) Where - is the mixture probabiity of two successive rays in the custer and, are the ray arriva rates Now whie discussing about the different environments through which a the RF signas pass, we can see that the radio signas are being refected from ta buidings, was, furniture and other conducting surfaces. We define different custers of different shapes and sizes for these incoming signas arriving at receiver at different times with different ampitude from different directions after mutipe refections. According to Saeh Vaenzuea mode the Power Deay Profie for our mode, was observed as the subsequent custers which were further attenuated in ampitude and aso arrivas within a singe custer decayed with time. These decaying patterns are exponentia within each custer which can be represented as:- E b, exp, (6) 697 www.ijergs.org Where is the integrated energy constant and is the inter- custer decay time constant of the th custer. Whereas it has been seen that the shape of the power deay profie can be different for the NLOS case of different environments so we can use this modified form:- b,.exp, rise.exp. rise E, (7) rise

Here, the parameter describes the attenuation of the first component, the parameter rise determines how fast the PDP increases to its oca maximum, and determines the decay at ate times. The custer decay time is ineary dependent on the arriva time of the custer, described as:- Where T 0 j (8) j describes the increase of the decay constant with the deay. The above parameters give a compete description of the Power Deay Profie of our mode. Apart from these parameters we wi be using some of the auxiiary parameters for the better comparison with the existing measurements RMS deay spread and mean excess deay are mutipath channe parameters that can be determined from a Power Deay Profie. The mean excess deay is the first moment of the power deay profie and it is described as: p p (9) Where p is the power measured at the time of the th ray arriva. The rms deay spread is the root mean square of the second moment of the power deay profie and is defined to be: (0) Where, p p Another auxiiary parameter such as the number of mutipath components that is within x db of the pea ampitude, or the number of MPCs that carries at east y % of the tota energy. These are determined from the power deay profie with the ampitude fading statistics. Naagami-m distribution is the generaized distribution which can mode different fading environments. It has much greater fexibiity and accuracy than Rayeigh, Rician or ogarithmic distributions. The foowing sma scae fading pdf is Naagami: m m m pdf ( x) x exp x ( m) () Where m / is the Naagami-m factor, (m) is the gamma function, is the mean-square vaue of ampitude. RESULTS Depending on the mathematica mode of Saeh- Vaenzuea mode and Naagami-m sma scae fading we generate MATLAB code for the channe mode condition. We characterized the environments in 8 different channe conditions. These channe conditions incudes a different environmenta conditions and frequency ranges. These conditions are basicay those environments which are common for signa transfer nowadays. 698 www.ijergs.org

We generated 00 impuse responses for the testing of our mathematica channe modeing. Each of these impuse responses were normaized to unit energy. These impuse responses were tested in a the environmenta conditions specified. Depending on the resuts observed we have dispayed a tabuar comparative study of a the channe conditions as foows:- CHANNEL NO. CHANNEL CONDITION MEAN DELAY(Ns) RMS DELAY(Ns) Number of paths within 0 db Number of paths capturing >85% energy Residentia LOS 6.0 7 5.3 54.6 Residentia NLOS 9.9 9 35.6 0.7 3 Office LOS 9.6 0 4.3.3 4 Office NLOS 8.4 3 30.4 45.3 5 Outdoor LOS 6.8 30 7.9 35.9 6 Outdoor NLOS 7.8 74 4.7 65 7 Industria LOS.6 9 6.7 8.7 8 Industria NLOS 3.9 0 8.5 86.6 Tabe No. CHANNEL NO. - RESIDENTIAL LOS Fig. - Impuse response reaizations This graph dispays the variation of 00 impuses with time. Different coours in graph denote different impuse reaizations. We can ceary infer from the graph, that impuse strength is higher in the time period between 0 to 50 (ns). Signa strength decays as the time increases and competey fades away after 00 (ns). Fig.- Excess deay 699 www.ijergs.org

This graph is for the reaization of excess deay with the different channe number responses. We can see from the graph that 68 th impuse response have the highest excess deay of 4.98 (ns) and 7 th impuse response have the owest excess deay of 3.5(nS). The red margin is set at the minimum excess deay required for our channe mode which is at 5.98 (ns). Fig.3- RMS deay This graph is for the reaization of RMS deay with the different channe number responses. We can see from the graph that 4 th impuse response have the highest RMS deay of 7.4 (ns) and 0 th impuse response have the owest RMS deay of 7.377 (ns). The margin at 6.57 (ns) is set according to the minimum required RMS deay for the mode. Fig.4- Number of significant paths within 0 db of pea Threshod energy for the channe is set at -0 or 0 db. We tried to find out the number of significant mutipath components whose energy is greater than 0 db pea. We compared the absoute vaue of st coumn of our obtained matrix with the threshod energy of a the impuse responses. For channe no tota no of significant mutipath components are found to be 5.3. 700 www.ijergs.org

Fig.5- Number f significant paths capturing > 85% energy To determine the number of significant paths capturing 85% of energy in channe, we compared the cumuative energy of a the received impuse responses with the sorted cumuative sum of energy of the individua impuses matrices. For channe number no. Of significant paths are 54.6. Fig.6- Average Power Decay Profie Power decay profie is exponentiay within each custer which can be easiy demonstrated with the observed graph. We have potted our graph between the average decay profie in (db) which is greater than the threshod eve of -40. Average power is highest within the range of 0-50(nS) and then it decays eventuay as deay increases. CONCLUSION Through this paper we have tried to mode the modified Saeh- Vaenzuea mode for the UWB channe. In this paper the mode describes the channe ony excuding the antenna effects. We have tested the designed channe in different environmenta conditions both in LOS and NLOS regions. The data sheet specified above gives the vaue of different parameter vaues of these channe conditions. We have discussed the graph observed for the channe no. i.e. residentia LOS in our resuts. The future scope of our project wi be the impementation of our channe in our practica Wi-Fi-ad system. The BER, PER and throughput of Wi-Fi system woud increase using this channe mode. REFERENCES: [] Moisch, A.F.; Cassioi, D.; Chia-Chin Chong; Emami, S.; Fort, A.; Kannan, B.; Kareda, J.; Kunisch, J.; Schantz, H.G.; Siwia, K.; Win, M.Z., "A Comprehensive Standardized Mode for Utrawideband Propagation Channes," Antennas and Propagation, IEEE Transactions on, vo.54, no., pp.35,366, Nov. 006 70 www.ijergs.org

[] Moisch, A.F., "Utra-Wide-Band Propagation Channes," Proceedings of the IEEE, vo.97, no., pp.353,37, Feb. 009 [3] Danies, R.C.; Murdoc, J.N.; Rappaport, T.S.; Heath, R.W., "60 GHz Wireess: Up Cose and Persona," Microwave Magazine, IEEE, vo., no.7, pp.44,50, Dec. 00 [4] Perahia, E.; Cordeiro, Caros; Minyoung Par; Yang, L.L., "IEEE 80.ad: Defining the Next Generation Muti-Gbps Wi-Fi," Consumer Communications and Networing Conference (CCNC), 00 7th IEEE, vo., no., pp.,5, 9- Jan. 00 [5] Mishra, Muesh K.; Sood,Netu; Sharma, Ajay K., Efficient BER Anaysis of OFDM System over Naagami-m Fading Channe, Internationa Journa of Advanced Science and Technoogy, Vo. 37, Dec. 0. [6] Chunei Yin; Guangjun Wen; Zhengyong Feng, "Simuation research of 80.n channe mode D in NS," Computer Science and Information Technoogy (ICCSIT), 00 3rd IEEE Internationa Conference on, vo.7, no., pp.530,533, 9- Juy 00 [7] Lansford, Jim; Stephens, A.; Nevo, R., "Wi-Fi (80.b) and Buetooth: enabing coexistence," Networ, IEEE, vo.5, no.5, pp.0,7, Sep/Oct 00 [8] Shuyan Jiang; Jie Peng; Zhi Lu; Junjie Jiao; Shanshan Jiang, "80.ad Key Performance Anaysis and Its Appication in Home Wireess Entertainment," Computationa Science and Engineering (CSE), 04 IEEE 7th Internationa Conference on, vo., no., pp.595,598, 9- Dec. 04 [9] Haiyan Li, "Channe Capacity and Channe Estimation of OFDM Utra-Wide-Band Systems," Computer Science and Eectronics Engineering (ICCSEE), 0 Internationa Conference on, vo., no., pp.5,56, 3-5 March 0 [0] Chavan, M.S; Chie, R.H; Sawant, S.R, Mutipath Fading Channe Modeing and Performance Comparision of Wireess Channe Modes, " Internationa Journa of Eectronics and Communication Engineering, Vo. 4, no., pp. 89-03, 0 [] S. Ghassemzadeh, L. Greenstein, T. Sveinsson, A. Kavcic, and V. Taroh, "Uwb indoor path oss mode for residentia and commercia environments, " in IEEE VTC 003- Fa, 003. [] A. Saeh and R. A. Vaenzuea, A statistica mode for indoor mutipath propagation, IEEE J. Seected Areas Comm., vo. 5, pp. 38 37, Feb. 987 70 www.ijergs.org