Wireless Innovation Forum Top Ten Most Wanted Innovations

Innovation #7: Context Aware Cognitive Radio

7.1 Executive Summary

Methods, tools, architectures, and languages need to be developed to enable cognitive radios to incorporate contextual reasoning into their decision processes. By adapting to dynamic contexts, cognitive radio algorithms can be automatically matched to changing conditions, wireless network performance can be significantly improved, end user experience enhanced, and network management time and costs reduced.

7.2 Applications

By incorporating context awareness into cognitive radio processes, the following will be feasible:

  • Automating communications leader communications resource management in a way that recognizes and facilitates reaction to the dynamic nature of major incident/disaster responses. This generalizes to disrupted context scenarios where the communications network was planned to accommodate assumptions (pre-defined contexts), but a radically different and rapidly evolving environment (contexts) is encountered.
  • Enabling smartphones to automatically adapt their wireless behavior based on what other users are nearby, what services the smartphone is trying to support, time-of-day, location, calendar information, and models of user intentions.
  • Improving spectrum sharing, management, and co-existence of cognitive radios by incorporating a better understanding of the changing capabilities, constraints, and goals of radio systems.
  • Increasing the market-size, applicability, and robustness of wireless devices by incorporating a context-aware software module (possibly with additional software defined radio capabilities) that adapts the behavior of the communications assets to the varying needs of different customers and changing situations
  • Simplifying the deployment and configuration of self-organizing networks (small-cells) in existing cellular bands and considered spectrum sharing bands (e.g., 3.5 GHz) by having radios automatically detect and monitor their operating context and adapting accordingly.

7.3 Description

Context refers to

  • information associated with a message that is not directly communicated with the message that influence the message’s meaning
  • a set of relationships that one object has with other objects.

While most cognitive radio literature considers the context inferable from a wireless chipset, a context-aware cognitive radio should have a wider scope of information to draw upon in its decision processes, including what the radio should be trying to do, what it is doing, what messages it is conveying, or the meaning of data in accessible databases.

Examples of valuable contextual information to cognitive radio processes include the following.

  • Locations, trajectories, and patterns of movement of radios and their users
  • Relationships between the users of radios
  • Characteristics of the radio users, such as their changing goals and objectives, calendars of activities, and general preferences
  • Services provided by and receiver characteristics of other radios and networks in the area
  • Current and intended activities and applications of the users of the radios
  • Meaning, and urgency of messages being carried over communications networks
  • External environment conditions, such as RF load density, general propagation and interference characteristics, weather conditions, time-of-day, and sensor data
  • Platform specifics such as power consumption, calculation and measurement precision, permissible operating environments and policies.
  • Metrics commonly used by cognitive radio designs such as RSS, BER, end-to-end delay, and packet loss rate.

Understanding the following three types of context will be particularly valuable to further development of cognitive radio applications.

Operational context
Operational context refers to the object relationships pertaining to the operation of the radio. This includes information such as what a user is trying to do, where the radio is located, the current needs of the situation, and the changing relationships with other users and radios. Understanding operational context is critical to automating communications resource management.

Communications context
Communications context refers to information associated with a message that is not directly communicated with the message that nonetheless influence the message’s meaning. Much operational context can be inferred from the meaning of messages conveyed by the communications system. Further, an understanding of communications context can lead to better cognitive radio decisions, new forms of encryption (by withholding context from unintended recipients), and more efficient communications (by understanding the context of intended recipients so that less contextual information needs to be transmitted).

Data context
Data context refers to the provenance, reliability, and relationships of data made accessible to cognitive radios, such as is envisioned for white space databases. This includes meta-dataabout the provenance and reliability of data, relationships with other data, and the objectives of the system. Understanding data context can help address Big-Data-like problems in the RF space, such as synthesizing and extrapolating meaningful patterns of behavior from accumulated sensor measurements.

  • As a new field, realizing context-aware cognitive radio will require the development or adaptation of several supporting technologies and innovations, including the following.
  • Extensible languages and tools for modeling and reasoning on contextual information drawn from a diverse and changing set of sources.
  • Mechanisms for automating the creation and updating of context models from observations and data shared from other sources and internal reasoning processes.
  • Algorithms for improved contextual pattern recognition to reduce ambiguity and rapidly determine the applicability of existing solutions.
  • Processes for innovating new solutions and interfacing with users when existing solutions are determined to be inappropriate to the determined context.
  • Programming languages (paradigms?) that allow end-users to easily specify how their communications assets should behave under varying scenarios.
  • Improved interfaces and processes for translating data (e.g., from sensors or online databases) into formats suitable for contextual reasoning
  • Methods and paradigms for interfacing with users to provide contextual information to the user and corrective feedback to the systems when ambiguous or erroneous contexts are identified
  • Big Data tools applicable to RF problems (i.e., Big RF) for collecting, correlating, weighting, and reasoning over disparate pieces of information in a timely, precise, verifiable, and accurate fashion at varying degrees of granularity
  • A flexible regulatory framework that will support the kinds of temporary, cooperative, and opportunistic applications that will emerge from context aware cognitive radio.