Path 3: Resource management

Description

The PU/SU concept can be generalized by introducing flexible access rights to all network resources on a local scale, not only spectrum. Nodes of a network, e.g., wireless devices, take their resources to the network. Resources such as memory and processing power are included in the shared resources. The applica-tions and interfaces of the wireless terminals are designed to make flexible use of available resources.

Wireless network resources can be classified into five types: 1) radio sources, including time, frequency, space, and power/energy; 2) built-in re-sources such as mass storage, batteries, and processing units, e.g., CPUs; 3) user interface resources, which refer to sensors as well as actuators integrated into wireless terminals like speakers, microphones, imaging devices; 4) social re-sources, which are considered part of a resource pool distributed across cooper-ating entities. Social resources are the individuals behind the wireless devices (e.g., controlling or taking decisions), as well as groups of them, such as in a social network. It is very important to consider these social resources in the overall picture of distributed resources, as individual and group decisions play a key role in the exploitation and management of distributed resources; and 5) connectivity resources, which typically include several cellular and short-range

air interfaces, can be seen as a fifth resource, as devices nowadays typically have more than one air interface. A large amount and diversity of distributed re-sources offer great potential for cooperation. Figure 2 has depicted the described resource classification, showing typical examples of these resources in wireless networks.

Users may play a key role by deciding which resources they are willing to share or, in some cases, the cognitive operation is embedded in the system and transparent to the user. In order to encourage nodes to share their resources, nodes need to obtain a clear benefit from this action. Incentive mechanisms in cooperation encourage cooperative behavior on a voluntary basis. For example, incentives can be better or cheaper services for the end-user and even enhance the e-reputation of a cooperative user. A rational entity chooses to cooperate as a result of an incentive mechanism that aims to bring benefits to the cooperating user and eventually all collaborating users.

Full awareness of all local resources, coexisting communication devices, etc.

and the ability to optimize communication according to the needs and capabili-ties of all players in the field open up possibilicapabili-ties: terminals or access points can act as relays, femtocells, local mesh structures, direct communication between terminals, and coordinated multipoint transmit and receive system (several an-tennas connected).

Research issues

The identification and management of resources in local networking environ-ment(s) must be extended to wireless access. This includes a minimum of the following broad research topics:

 Resource description: Before resources can be shared they need to be defined and described efficiently. Devices have to agree on a common language to describe both the resources and the need for them just to make resource sharing possible.

 Resource discovery: Resource discovery is the process of locating dis-tributed resources on a network. Devices can formulate their needs and publish their resources using resource description protocols. Devices can locate the necessary resources (or services) using resource discovery protocols. In addition, real measurements can be carried out to assess the status of given resources, particularly physical resources.

 Learning: Learning is a crucial part of a cognitive network and makes the operation of CRs more efficient than in the case in which it is only possi-ble to have information availapossi-ble at design time. A cognitive network should be able to learn from the past and present results of its actions and from the actions of PUs in order to improve its future performance.

 Coordination systems must be employed for resource access control in order to share resources efficiently among wireless end-users. Coordina-tion systems allow devices to use other devices’ resources or permit pooling and scheduling of resources among those nodes requiring ser-vices. Coordination is a great challenge in a mobile wireless network because of its distributed and highly dynamic characteristics.

 A clearing mechanism is needed to permit transactions, generally with an authentication or authorization process. It may also include payment to obtain permission. In several peer-to-peer networks, mutual resource sharing is incorporated into the clearing mechanism. Peers have to con-tribute their own resources in order to be cleared to use the resources of other peers. The level of the received service improves with increased sharing of the users’ own resources. To enable a trust and clearing me-chanism for cognitive spectrum use, some kind of authorization may be needed for CR devices. They should be tested carefully to meet tight in-terference requirements, and only devices proven to work properly would be cleared to coexist with PUs.

It may be challenging to convince people to share resources, for privacy and security reasons. In fact, security, privacy, and trust issues of wireless sharing of a network are big concerns for people. Thus, in some cases the end-user may decide with whom to cooperate and what resources to share. User decisions on what and how to share the resources of their devices are part of the cooperative strategies that need to be devised.

Beneficiaries and use cases

Network operators can use resources more efficiently and support more users with new services. For example, if the network is heavily congested, the user’s wireless device senses the congestion in the network and asks if another user is able to tolerate some delay. If this user can tolerate some latency (e.g., because his/her traffic is not delay sensitive), he/she may accept sharing his/her resources

to help other users to improve their QoS. It is also possible to borrow resources from other participants in situations in which the user’s own resources are not powerful enough for a given task. For example, it may be the case that a given user does not have enough signal strength in his/her cell phone and that the situation of a close-by user is much more favorable, as he/she uses a different service provider.

By sharing all resources cooperatively, even a user with an unfavorable radio channel can obtain the required resource through his/her neighbor device (e.g., air-interface or operator sharing). A simple example of a service provided by a wireless grid is cooperative audio recording in which the stereo sound is pro-duced by combining recordings made by multiple simple devices. Cooperative media downloading is also a possible application scenario for a grid. If several mobile users want to download the same media from a certain source, they can download different parts of the media over a cellular link and then exchange the parts over short-range links, which is a more energy- and spectrum-efficient way than downloading all the data from the source individually. Operators can use resources more efficiently and support more users with new services.

Potential use cases include in-home communications in which local clusters are formed to allow low power transmissions. CR will be used for intelligent home services. Control systems are different, but intelligent systems take care of most of the configurations. There is no need to worry about different radio inter-faces or cable connectors.