Article Preview
Top1. Introduction
Cognitive Radio (CR) networks (Akyildiz et. al, 2006) enable for the efficient exploitation of radio spectrum parts to deploy emerging mobile computing and wireless networking architectures. CR technological solutions consist of nodes that are able to change their transmission settings, according to the available radio spectrum at local level. Such nodes support the capability for sensing large parts of the radio spectrum, by dynamically using locally un-exploited frequencies. This capability enables for the proper design and development of novel wireless networking architectures, based on new policies for the opportunistic access of the available radio spectrum in specific geographical regions, such as the “television white spaces” (i.e. TVWS) (Bourdena et. al, 2014). TVWS consists of television/broadcasting channels that are available after the digital switchover process or are totally un-exploited due to frequency planning issues (i.e. Interleaved Radio Spectrum) (Bourdena et. al, 2011, December). Hence, the deployment of CR networking architectures, operating over the TVWS, is highly related with the radio spectrum management models that will be exploited (Bourdena et. al, 2012, October; Bourdena et. al, 2012, December) in emerging communication systems. In this direction, “Spectrum of Commons” model can be adopted in ad-hoc cognitive radio networking architectures, where the allocation of the available resources is performed at local level by the nodes, instead of exploiting a centralized unit, like a spectrum broker (Hossain et. al, 2009). This dynamic radio spectrum access by the CR nodes causes new challenges in the design of novel networking protocols at different layers. The design and development of effective routing schemes, is an important process for the proper operation of such emerging computing environments. The ad-hoc CR networks are based on self-configuring architectures (Hossain et. al, 2009), where the routing process is challenging, in comparison with the routing schemes adopted so far in conventional wireless networks. The main difference is that the radio spectrum availability in CR networks is highly related with the presence of primary nodes (e.g. television services broadcasting sites), making challenging the use of a Common Control Channel (i.e. CCC), to create and maintain a stable route among secondary CR nodes.
On the other hand, the energy conservation issues are crucial for the efficient deployment of future ad-hoc CR networks. The energy conservation schemes that will be adopted have to be reactive, in order to tune the energy levels of the nodes based on estimated parameters (e.g. traffic, capacity (Mavromoustakis, 2012, May)). In addition, the energy-efficient schemes have to consider the bounded end-to-end delays of the data transferred among the nodes. Since the network lifetime is strictly associated with the transmission characteristics that are exploited by the source node to transmit data to a destination node, as well as the routing protocol used (Shpungin, 2011, June), a scheme that associates the temporal traffic-aware behaviour of the nodes (Mavromoustakis et. al, 2010, September) in an end-to-end path is crucial to be studied. In (Charalambous et. al, 2012, June), the sleep-proxy nodes calculate the interval of the nodes activity periods, based on the capacity and the estimated inter-cluster total energy consumed within a time frame. The authors of the previous scheme applied the traffic model, as well as evaluated the traffic volume characteristics for a specific time window frame but not in CR ad-hoc network architectures.