Cognitive Radio Infrastructure using Spectrum Observatory and Layered Detection Strategy for efficient detection. Kunal Rele Abstract: Cognitive radio is an amazing technology that allows low cost voice and data services by identifying opportunities in spectrum, space, code and time. This opportunity detection and adaptability requires infrastructure planning for accurate and fast detection of its environment. It also requires the CR to readjust its spectrum when a licensed user becomes active. An infrastructure is proposed for the Cognitive radio network which will have dedicated or Base Station integrated Spectrum Observatories (SO). They will be used to sense the macro environment. The cognitive radios which will be implemented using Software defined radios will have GPS capabilities and will communicate its sensed environment information with the base station or a master. The cognitive radio will use OFDMA technology for communication but is not limited to it. It can also use WiMAX and LTE when they are rolled of in the market. The Master/BS will build a matrix of space, time, and frequency and use an algorithm to efficiently maximize the utilization.

I. Introduction The FCC’s decision to allow secondary intelligent device in the TV white space has given a boost to develop highly flexible devices for mobile broadband. White space is the term used by the FCC for unused TV spectrum. This VHF and UHF spectrum provides superior propagation and building penetration compared to other unlicensed spectrum in other bands like the 2.4 and 5 GHz bands. With the advent of iPhone’s the Mobile IP traffic has gone up considerably with the data transfers reaching 100 PetaBytes in the first quarter of 2009. There was a concentration of effort to increase the wired backbone of US networks and it has reached a considerable proportions. The bottleneck now is the mobile wireless connection. Due to the static, licensed nature

of the mobile broadband the providers have to put a limit on the capacity or increase their rates or use some other dynamics to handle the scarce resource. But with the introduction of TV white pace this is all about to change. Cognitive radio devices are an ideal candidate for this particular band. The can be trained to sense the environment and adapt their transmission accordingly. It is true that the spectrum is scarce but it also true that it is currently not used efficiently because of its static allocations. The cognitive radio networks will change that because they will be able to sense the wireless environment and transmits in the spectral, temporal and spatial holes that are present in the wireless environment. These holes were present because initially we didn’t have the capabilities to build such intelligent radios and hence we assigned static channels to allow for protection from each other. But with the advent of this new technology there are some very stringent rules that the FCC has applied to the Cognitive Radio devices that will work in these bands. There are some new challenges in network planning that one has to take into consideration for this new technology. Location awareness and spectrum sensing is laid out as an important criteria for these cognitive radio networks. Fixed devices are permitted to transmit up to 30 dBm (1 watt) with up to 6 dBi antenna gain, while portable devices are permitted to transmit up to 20 dBm (100 mw) with no antenna gain. Television broadcast signals are protected with a protection contour. Protection contour is the distances around the TV transmitter

within which the cognitive radio network cannot operate. Fixed Cognitive radios are not permitted to operate on channels adjacent to the TV broadcast channel. Portable devices are permitted to operate on an adjacent TV channel but with maximum allowed transmission power of 16 dBm.

environment fast and accurately. This is achieved by distributed sensing and cooperative updates. The sensing itself works in layers and uses elimination to minimize the work done by each component as well as decreases the time needed to do the sensing.

The first interference avoidance mechanism that a cognitive radio uses is the location awareness or it should have geolocation capabilities. Information about licensed transmission in the various TV channels is stored in a database that can be accessed using internet access. These licensed transmissions include ATSC (Digital TV) highpower broadcasts, ATSC and NTSC (Analog TV) low-power transmitter, and wireless microphones used by the broadcast industry.

It also describes the implementation of a cognitive radio networks which give a deeper implementation of the issues and possible solutions.

The second interference avoidance mechanism is spectrum sensing in which the cognitive radio networks must observe the various frequency channels and determine if these channels are occupied by any licensed transmission. Cognitive radios should be able to detect the presence of (digital and analog) TV signals and wireless microphone signals at a received power level −114 dBm received power level as well as other cognitive radio networks since these can be heterogeneously implemented. The -114 dBm power level translates into a SNR of around −15 dB. There is still the competing camps on sensing versus not sensing, what are the appropriate sensing levels and who should be protected or not, but there appears to be general consensus that the geolocation database should be secured and should permit more frequent updates and accesses. This paper discusses some solutions to the problems involves in sensing the

II. Infrastructure The cognitive radio network describes here work in infrastructure environment. The networks consist of master/Base station slave/Client equipments configuration. This particular infrastructure targets the problem of hidden node through distributed sensing. It decreases the sensing time and increasing the accuracy by introducing co-operation between different networks as well as having different layers of sensing.

Fig. 1 Spectrogram of UHF TV bands in Chicago over a week Figure 1 shows the spectrogram of TV UHF bands in Chicago. The spectrogram is of the data taken over a week period. The high

powered TV transmission can be seen as bright red bands and the unoccupied spectrum, the white space is shown by dark blue bands. There are some light blue bands that are weak TV transmission. These bands have to be sensed and avoided too. As can be seen there is a lot of spectrum that is not used. Cognitive radios can make use of these holes in the spectrum for unlicensed transmission. Fig. 3 Transmission period and quite periods in the cognitive radio network Figure 3 shows that we will have quite periods between transmission periods in the cognitive radio network. All the cognitive radio’s will go quite and start sensing during this period. The purpose of these quite periods is two folds. First, obviously we need t sense the environment periodically since its changing very fast. Secondly, it brings the noise floor down so that we can sense the low powered licensed signals efficiently. A. Macro sensing Fig. 2 Cognitive radio infrastructure with two layered detection Figure 2 shows an infrastructure that is proposed for the cognitive radio implementation in the TV bands. It consists of Spectrum Observatories (SO) that can be independent or can be integrated with the Base stations of the cognitive radio network. This particular implementation can be implemented for WRAN’s, WWAN’s, WLAN’s. Figure 2 also shows a WLAN network in which we have a master or a base station that communicates with the mobile or fixed nodes.

As shown in figure 2, the master or the base station of the WLAN network communicates with the SO to get spectrum occupancy information for high powered signals as well as all the licensed signals. The geo-location database is integrated in the spectrum observatories. The cognitive radio network master also updates the channels that it allocates to its nodes to the SO. So when there is another CR network in the neighborhood it gets the information without the need to sense. This is called cooperation and it decreases the time required for detection and avoidance.

Adjacent SO’s communicate with each other using Inter-SO communication. They exchange their sensed information of high powered signals, triangulate the unlicensed transmitters and create a 3D view of the wireless environment. B. Micro sensing Macro sensing is good at detecting high powered signals. The cognitive radio networks themselves can use low powered transmissions and they need to detect and avoid each other efficiently too. As mentioned earlier these cognitive radio networks can be heterogeneous in various respects including power. The micro sensing would include 2 layers of detection as well, power and device detection. The Macro sensing step has minimized the amount of work we need to do in micro step. This is called efficiency by elimination. We are distributing the work between SO, and the Cognitive radio nodes. The power detection will eliminate the amount of work we need to do in the device detection step which is more advanced and time consuming. 1. Power Detection Power detection as shown in figure 4 includes edge detection technique to isolate the occupied bands efficiently. It includes noise filtering and derivatives of sample points to detect the starting edge and the ending edge of a signal. The starting edge is represented by the positive peak of the first derivative plot and the end is represented by the negative peak. The height of the peaks denotes the power level of the signal.

Fig. 4 Edge detection, primary micro detection scheme for white space detection All the nodes do this detection in a distributed manner. This data is send to the master for data fusion and efficient allocation of channels to each node. The master node sends its channel allocation scheme information to the nearest SO. 2. Device detection In this step the cognitive radio nodes actually identify the devices that are working in the vicinity. This step is required the environment is very crowded and the cognitive radio network needs to use efficiently utilize the temporal and spatial holes. It is also required to identify heterogeneous services when the power level from its current base station starts decreasing. As the cognitive radios are implemented using software defined radio technology they will also be able to adapt the transmission depending on the service technology available. Our particular cognitive radios will use the Software Communication Architecture (SCA) for the

software radio implementation. This technology includes waveforms constructed using XML files. They can be pre-cached in waveform cache and then be written to DSP to make the transition from one technology to other fast. For example if it’s currently attached to a Base station that is using WiMAX and it detects there is a LTE service available then it can starts precaching the LTE waveforms and then writing it to the DSP when it handovers from WiMAX base station to LTE base station.

is fairly low powered. Because of these characteristics the detection and avoidance of wireless microphones become very difficult. Hence, there are some proposals like registering the band with the central database, allocation 2 channels specifically to wireless microphones and using high powered beacons for easy detection. III. Cognitive Radio Features The feature chosen for the implementations are in accordance with the rules stated by the FCC. In the 6 MHz channels adjacent to the operating channel, emissions from Cognitive radios will be at least 55 dB below the highest average power in the band. For fixed we will use a maximum transmit power of 30 dBm and an antenna gain of 6 dBi. For mobile we will use maximum transmit power of 20 dBm and a 0 dBi antenna gain. In both cases we will use a signal bandwidth of 5 MHz, to allow for the guard band.

Fig. 5 Simulated ATSC channels ATSC TV signals as shown in figure 5 are modulated using 8-level vestigial sideband modulation (8VSB) occupying 6 MHz channel, with a pilot carrier which contains approximately 7% of the total signal power and is located at approximately 310 kHz above the lower edge of the channel. Various feature of ATSC waveform like the shape of the waveform, the pilot, cyclostationary features etc can be used for detection. Wireless microphones have maximum bandwidth of 200 kHz. Therefore, multiple such devices may operate within a single TV channel and their carrier frequency locations are not fixed. There is no specific requirement on the modulation scheme for wireless microphones. Also the transmission

Fig. 6 Comblock for generating ATSC and WiMAX signals

that my Cognitive radio will change waveform (E.g. WiMAX to LTE) during handoff), the sensing delays, Throughput, packet drops during handoff.

Fig. 7 WiMAX signal transmitted using Comblock The Master/Base station is the main element in the network because it decides which channels get allocated to each node. It can use two types of implementation. When it has to coexist with another network in a heterogeneous manner it uses MultichannelDynamic Frequency Selection (M-DFS), in which it assigns channels to the nodes that have least interference. It assigns this subset of the available channels in a manner that minimize the overall outage probability and improves system throughput. Figure 6 shows Comblock which is a device that can be used to transmit time sampled values that are stored in its flash memory. The samples are created using simulation in Matlab. We can test our scheme by creating future environment using the fake transmissions from this device. Figure 7 shows a simulated Cognitive radio transmission which can be transmitted using Comblock to check interoperability and isolate the bottlenecks. Simulation As mobility and heterogeneous networks are the main aspects in my proposed model, I am testing handoff delays, handoff delay variations (Jitter) (It should be remembered

Fig. 7 Internetworked Architecturea [a. Courtesy: Munasinghe, K.S.; Jamalipour, A., "An architecture for mobility management in interworked 3G cellular and WiMAX Networks," Wireless Telecommunications Symposium, 2008. WTS 2008 , vol., no., pp.291-297, 24-26 April 2008]

Figure 7 shows a heterogeneous architecture model. My model will be similar but will also include LTE

Fig. 8 OPNET Simulation model1 Figure 8 shows the OPNET simulation model and environment. This model includes certain features that are currently not present in the simulation models. We will be doing simulation using actual experimental data and plugging them into the model. The delay for changing the waveform from WiMAX to LTE will be noted using actual experiments. The backend abstraction layer will use mobile IP

for managing mobility, Session Initiation Protocol (SIP) for real time session negotiation and management. Already research models will be used to simulate the dynamics caused by MIP and SIP characteristics. The delay in sensing and location awareness will be experimentally calculated at different points in the architecture and their interactions will be integrated in the simulation models. In addition to above mentioned parameters End-to-End delays over WiMAX and LTE interfaces will be noted. Conclusion Mobile broadband access is growing at a tremendous rate. With the advent of iPhones, Android the Mobile IP traffic is shooting to the sky. Mobile TV and Virtual presence are around the corner too. They are waiting for a revolution in the wireless industry that would allow this kind of application to take off. But the current wireless networks are becoming bottlenecks because of their static allocation schemes. Hence Cognitive radios with its Dynamic Allocation schemes now allow us to utilize the opportunities that were not present earlier. The main challenges like sensing and adapting to the environment fast, seamless connectivity in a heterogeneous network and high mobility are studied in this paper. The solutions to these problems are also proposed like layered, distributed sensing with cooperation. SCA SDR architecture in a dynamic environment is proposed. Simulation to test the heterogeneous handoff, mobile IP, seamless session between heterogeneous networks is discussed. References [1] Stephen J. Shellhammer, Ahmed K. Sadek and Wenyi Zhang., Technical Challenges for Cognitive Radio in the TV White Space Spectrum., Corporate Research and Development.

[2] McHenry, M., “NSF spectrum occupancy measurements project summary,” Shared Spectrum Company, Vienna, VA, Tech. Rep., August 2005. [Online]. Available: http://www.sharedspectrum.com/?section=measurements [3] Mitola J.,Maguire, G., “Cognitive radio: Making software radios more personal,” IEEE Personal Commun. Mag., vol. 6, no. 4, pp. 13–18, Aug. 1999 [4] Jondral, F.K., "Cognitive Radio: A Communications Engineering View," Wireless Communications, IEEE , vol.14, no.4, pp.2833, August 2007 [5] Haykin, S., "Cognitive radio: brain-empowered wireless communications," Selected Areas in Communications, IEEE Journal on , vol.23, no.2, pp. 201-220, Feb. 2005 [6] Jondral F.K.: “Software-defined radio-basics and evolution to cognitive radio”, Eurasip J. Wireless Communication Network., pp. 275–283, 2005 [7] Ghasemi, A.; Sousa, E.S., "Spectrum sensing in cognitive radio networks: requirements, challenges and design trade-offs," Communications Magazine, IEEE , vol.46, no.4, pp.32-39, April 2008 [8] Cabric, D., Tkachenko A., Brodersen R., “Experimental study of spectrum sensing based on energy detection and network cooperation,” in Proc. ACM Int. Workshop on Technology and Policy for Accessing Spectrum (TAPAS), Boston, MA, Aug. 2006 [9] Kyouwoong Kim; Akbar, I.A.; Bae, K.K.; Jung- sun Urn; Spooner, C.M.; Reed, J.H., "Cyclostationary Approaches to Signal Detection and Classification in Cognitive Radio," +ew Frontiers in Dynamic Spectrum Access +etworks, 2007. DySPA+ 2007. 2nd IEEE International Symposium on, vol., no., pp.212-215, 17-20 April 2007 [10] Taher, T.M.; Al-Banna, A.Z.; Ucci, D.R.; LoCicero, J.L., "Characterization of an Unintentional Wi-Fi Interference Device - the Residential Microwave Oven," Military Communications Conference, 2006. MILCOM 2006. IEEE, vol., no., pp.1-7, 23-25 Oct. 2006 [11] Taher, Tanim M.; Rele, Kunal; Roberson, Dennis, "Development and Quantitative Analysis of an Adaptive Scheme for Bluetooth and Wi-Fi Co-Existence," Consumer Communications and +etworking Conference, 2009. CC+C 2009. 6th IEEE, vol., no., pp.1-2, 10-13 Jan. 2009 [11] Munasinghe, K.S.; Jamalipour, A., "An architecture for mobility management in interworked 3G cellular and WiMAX Networks," Wireless Telecommunications Symposium, 2008. WTS 2008 , vol., no., pp.291-297, 24-26 April 2008