Our last Tech Topic introduced the technologies of software defined radios (SDR) that allow for dynamic and adaptive software implementations of radio functionality. With SDR as an implementation strategy, the concept of cognitive radio (CR) technology was then introduced. The potential inherent in CR technology has given rise to rapidly emerging radio developments being pursued by the defense and commercial communities with the promise of inexpensive adaptable radio architectures in the future.1 CR devices can sense, detect, and monitor the surrounding radio frequency (RF) conditions including interference, assess frequency availability and reconfigure their own operating characteristics to best match those conditions. A cognitive capability that can make real-time autonomous decisions for radio operations can increase spectrum efficiency leading to higher bandwidth services as well as reduce the burdens of centralized spectrum management by public safety communications officials.
The flexibility and rapid growth of SDR implementations is apparent in two recently released equipment packages that capitalize on SDR technologies. The first is from Thales Communications, Inc., and is called the Liberty radio.2 Designed for the homeland security segment, this radio is intended to operate in the public safety frequency bands 136-174 MHz, 380-520 MHz, 700 MHz, and 800 MHz and operate in several modes including Project 25 (P25), P25 trunked, and conventional frequency modulation (FM).
The second piece of new equipment is the RF-1033M radio from Harris Corporation.3 This radio is designed to operate in the VHF low (30-50 MHz) and high (136-174 MHz) bands as well as the UHF (380-512 MHz) band. It is intended to operate in P25 modes and analog AM and FM modes.
These software-based radios represent only a first step, however, on the road to true cognitive technologies. Although these two radios are reconfigurable, they are not completely compatible with each other because they do not have the same operating frequency bands and operational modes. Accordingly, there remains much potential for dynamically reconfigurable radios but much work still needs to done to achieve true interoperability. So what advantage does reconfigurability provide?
As noted in our last topic, the DARPA XG and WNAN Programs are investigating much more dynamic frequency selective radios based on agile radio architectures that can respond dynamically to the radio's surrounding environment. The real value of this type of radio and network architecture is the efficiency of operation in establishing an interoperability environment over a flexible frequency range and over reconfigurable operating modes. Based on an assessment of their operating environment, that may also include an evaluation of location identification information and any particular operating rule set, i.e. a "policy-based" rule set. In this manner, these devices can modify their operational parameters such as frequency, modulation schemes, and transmit power, in order to capitalize on available spectrum or other resources.
Perhaps the key to the success and the future development of cognitive radio lies in the ability of developers and practitioners, that is, the first responder community, to establish the policy rule set by which the radios will operate - and this harkens back to Tech Topic #1 and to the fundamental premise of interoperability. There must be a priori agreement at the HUMAN LEVEL to establish the rule set for the operation at the MACHINE LEVEL. In this regard, several issues form the basis for a policy rule set - a few of these are addressed below.
- Network affiliation. More than any other factor in prearranged network operations is agreement on who will be allowed to operate in the radio network. This is not a factor for negotiation by the radios themselves! It must be accomplished by the radio users who will establish the authentication and verification policies and rules for the radio set.
- Frequency bands of operation. Although there is substantial promise in CR to capitalize on agile frequency use and the ability to use multiple frequency bands concurrently, at present the rules as outlined by the Federal Communications Commission are designed to allocate specific frequency bands for specific uses. Therefore, for the time being, CR must operate in assigned "lanes" (not to use a pun, but...) until such time as the technology matures to the point that flexible use frequency assignments and operating rules are possible. This means that agreement must be reached in advance on the exact frequency bands to be utilized. The promise of dynamic frequency assignment over large potential bandwidths is there (the spectrum efficiency factor!); but for the moment, it is a long way down the road and there must be sufficient confidence that the technology will support the requested service while at the same time protecting other legacy technologies and frequency allocations. The potential for a paradigm shift in frequency allocation to create a "pool" of CR frequency bands, as suggested by Nancy Jesuale and Bernard C. Eydt in Mission Critical Communications Magazine, is there, but a lot of agreement needs to be accomplished first.4
- Operating Modes and Parameters. With SDR technology as the means to implement a dynamically reconfigurable CR, there must be a priori agreement on the applications, program interfaces, and transmission parameters; that is to say, there must be agreement on the physical layer characteristics of transmission. The true glory of CR may be in its inherent ability for radios to negotiate these parameters; nevertheless, there must some a priori set of rules that forms the basis for negotiation and decision making. Current efforts at dynamic frequency selectivity and dynamic power control are only rudimentary. There must be policy initiatives for all of the transmission protocols including bandwidth, modulation scheme, channel assignment, forward error correction algorithms, security protocols, just to name a few. But much more exciting options lie in the future, such as location-based alternatives that rest on geolocation determination used in conjunction with a radio operational rule set.
- Proprietary Vendor Issues. At present, equipment vendors are moving rapidly to develop SDR and CR-based radios. Nevertheless, they are still built on proprietary foundations that remain the domain of the manufacturer. In this regard, FCC rules require original manufactures to have responsibility for software configuration control for software security purposes. But as long as vendors are insistent on their own proprietary SDR implementations, there is little chance of a priori agreement on a policy rule set for any large scale CR implementation. At the same time, as long as vendors license their software (for profit purposes) for others to use, there may be less incentive for a common policy-radio framework.
- Legacy Radio Evolution. CR is a technology of the future that provides for dynamic reconfiguration of the radio's operational characteristics. This is a positive trait for maintaining interoperability with legacy radios; an imperative for constituent agencies that are dependent on intermittent funding for improvements. However, the cost of CR must be such that a natural evolution can occur as legacy radios come to the end of their normal life cycles. At some point, the reverse compatibility issue becomes a hindrance to the development of new capabilities and a line of demarcation must be established just as the evolution of software packages must do.
- Operational and Technical Control. Without question, the key factor in the success of any CR network will be the ability of the controlling elements to establish and maintain control - administratively and technically - over the operation of the network. This includes not only the technical aspects of the policy rules of the network, but also includes the overarching control framework that determines who may roam into the network, what the quality of service criterion will be for all users, and other "cost of doing business" factors.
These are a few of the myriad of issues that must be resolved before cognitive radio will be able to meet its potential. The first hints are now obvious though that CR can meet the expectations. In our next topic, we'll take a look at how license exempt spectrum use, one aspect of CR, is already benefiting the public safety community.
1 The Defense Advanced Research Projects Agency (DARPA) XG Program has evolved and morphed into the DARPA Wireless Network After Next (WNAN) Program that is intended to create a flexible and adaptable radio architecture for self-forming ad hoc military networks. See http://www.darpa.mil/sto/smallunitops/xg.html.
4 See Jesuale, Nancy and Eydt, Bernard C. "Spectrum Paradigm Shift", RadioResource Mission Critical Communications Magazine, Volume 23, No. 3, March 2008, pp. 83-91.