Dynamic Spectrum Access for Wireless Networks

Wireless communication is the fastest growing segment of the communication industry. With the successful deployment of cellular networks in licensed bands and Wi-Fi networks in unlicensed bands, users have anytime, anywhere connectivity with the networked systems leading to the Internet of Things (IoT). Traditional wireless networks rely on static spectrum assignment where the government regulatory bodies, such as the Federal Communication Commission (FCC) in the United States, assign the Radio Frequency (RF) spectrum to the service providers in an exclusive manner for long term and vast geographic area. Most of the usable RF spectrum are already assigned to certain services leaving no bands for further development of new wireless systems. Furthermore, when everything (such as refrigerator, microwave oven, smart car, etc) is connected to internet, this scarcity would be more severe. However, recent studies show that the static RF spectrum assignment leads to inefficient use of RF spectrum since most of the channels are used only from 15 to 85 % or idle most of the time [2, 12, 14]. Thus the bottleneck created is not because of lack of RF spectrum but because of wasteful static assignments of RF spectrum for long-time and vast geographic area. In this chapter, we present an overview of cognitive radio network, spectrum sensing techniques, and spectrum access methods.

In spectrum overlay approach, SUs coexists with PUs and share their spectrum. However, each SU has imposed transmit power constraint so as not to create any harmful interference to active PUs. The main goal of each SU is to maintain lower interference level than the specified tolerable level at PUs while sharing PUs’ licensed bands dynamically. To maintain low interference level, SUs must transmit with lower transmit power. Thus power control is essential for each SUs. In wireless communications, power control is performed to satisfy the specified minimum signal-to-interference-plus-noise (SINR) to get desired data rate [2, 7]. For voice communications, once target SINR level is met, there will be no improvement in voice quality by increasing power or SINR. However, for data communications, increase in SINR results in increase in data rate. Thus, SUs try to increases their SINR values by increasing their transmission powers while satisfying their imposed power constraints in data communications.

Opportunistic spectrum access (OSA) is an emerging concept for spectrum overlay based spectrum sharing model where SUs identify spectrum opportunities in licensed bands and use them opportunistically without interfering with PUs [12, 20, 34, 36]. In spectrum overlay approach, as SUs are not allowed to co-exists with PUs in the same channel, they are required to either sense spectrum to find idle channels or search for idle channels in spectrum database [1, 8, 27]. Spectrum database can maintain geolocations of idle bands and provide a global view on entire frequencies which could be used to find best suitable channels for the SUs. Based on the global view of wideband RF regime, SUs could be granted a channel that has more adjacent channels so that SUs could implement channel bonding for higher data rates [13]. To prepare and update the spectrum database, database server can get spectrum occupancy information from PUs’ infrastructure (e.g. base stations or access points) in a real-time basis. Alternatively, spectrum sensors (e.g., crowd sourcing for sensing [9, 38]) could be deployed to collect information about channel status. Based on the collected information, spectrum server can process data (with the help of cloud computing platform) to create a spectrum maps for spectrum opportunities for different wireless networks such as satellite, WiMAX, Wi-Fi, cellular, TV, etc. Spectrum servers could be associated with a single or multiple spectrum brokers. When a SU wants to access spectrum opportunities, it searches the geolocation database for idle bands. If there is an idle band available in given location and time, SU would access it. Otherwise, the SU has to wait until it finds spectrum opportunities that meets its needs.

This chapter presents cloud integrated dynamic spectrum access in cognitive radio networks where most of the computing and storing function are performed using data offloading to cloud computing platform. The SUs are considerably constrained by their limited power, memory and computational capacity when they have to make decision about spectrum sensing for wide RF band regime and dynamic spectrum access. The SUs in CRN have the potential to mitigate these constraints by leveraging the vast storage and computational capacity of cloud computing platform [19, 21]. Specifically, cloud computing based dynamic spectrum access has following advantages: (a) Power saving in mobile devices: As SUs search the spectrum opportunities in the geolocation database, power needed for SUs to sense the RF spectrum for wide range of bands will be saved. By leveraging the cloud computing and storage resources, SU mobile devices can extend their battery lifetime; (b) No harmful interference to PUs: Chances of mis-detection of spectrum opportunities can be significantly reduced when SUs are required to search the database instead of sensing and identifying spectrum opportunities by themselves. Furthermore, the aggressive SUs can be monitored and possibly penalized by incorporating a cloud assisted manager to oversee the overall system; (c) Compliance with the requirements of the regulatory body: Recently the FCC in the U.S. [1, 12] mandates that the SUs must search geolocation database for spectrum bands instead of sensing and identifying the spectrum opportunities themselves. Thus, the proposed approach follows the recent proposal by FCC and can be implemented easily in real systems; and (d) Outsource computing on mobile devices: Typical mobile devices used by SUs have limited computing capabilities that limits the scalability of cognitive networks. Using proposed approach, the computation performance of SUs is significantly enhanced by outsourcing the streaming computation tasks to the cloud computing systems.

Vehicular communication networks are expected to provide safety and comfort services to passengers and drivers. To support diverse set of services and applications, vehicular network is expected to used variety of wireless access technologies for Vehicle-to-Vehicle (V2V) and Vehicle-to-roadside (V2R) communications. V2R communication could introduce high delay as roadside communication unit relays the information from source vehicle to destinations vehicles. Thus, for time critical messages, V2R based communication is not suitable to notify drivers in a timely manner. V2V based communication in VANET is performed in a peer-to-peer basis and the intended vehicles could exchange their information directly using single hop or multi-hop communications. In this case, performance of VANET depends on connectivity among vehicles since reliable connectivity for single hop or multi-hop communication is very important to forward time-critical information. The connectivity in VANET is directly related to density of vehicles, relative speed of the vehicles, association time of wireless technology, and transmission range and frequency bands used by vehicles. In this chapter, we present cloud assisted cognitive radio enabled vehicular communications.