ALOE Concept and Motivation

The ALOE concept tries to address the main issues related with the Software Radio technology. It assumes the need of supporting, partial or total, reconfiguration of the radio processing chain, the need to deploy the radio application onto a heterogeneous hardware platform capable to assume the high computing requirements of modern wireless systems, the need to assure strong real-time execution constraints, the need to easily develop and integrate software modules and the possibility to execute them in different types of processors assuming an easy portability.

One of the most relevant objectives in the process of defining a common framework to develop and deploy software radio applications is to eliminate platform (hardware and support software) dependencies. On the other hand, radio applications are built through a set of software modules (hardware or software based) that communicates among them. The common used terms denominates such modules as “objects”. Therefore each one of such objects are signal processing blocks, some of them with important requirements in terms of computing resources, that needs to acquire/deliver information from/to other objects through proper interfaces. Nevertheless, one of the most relevant assumptions of PHAL-OE is that such interfaces between objects are unknown at object development time. Only at execution time all the required objects will be integrated and the entire radio application will be built.

All these objectives have pushed us to define and develop a multi-platform software abstraction layer and execution environment, ALOE (Abstraction Layer & Operating Environment), capable to provide such features. The list of the main functionalities where the PHAL-OE must provide support on includes:

• Flexibility. The framework must efficiently implement the flexibility concept required by SDR. It is assumed to be based in the capacity to facilitate the reconfiguration as the basic mechanism providing flexibility.

• Execution control management. The coordinated execution of the whole system must be assured.

• Hide the platform heterogeneity to the radio application. Abstraction layers are required are basic mechanisms to provide such feature. • Computing resource management. Under a scenario with limited computing resources the need of a specific mechanism to manage the available resources promotes the efficiency in the whole system and extends its interactions with the radio resources management part. In addition it is capable to assure the overcoming of the real-time constraints.

• Computing objects data packet oriented messaging. Not processor (device) oriented communication mechanisms. Therefore a data packet oriented communication network among the heterogeneous processors is built.

• Parameter or variable (signals) evolution capture during the execution of the application.

• Processing resource parameters evolution captured for autonomous management or control.

• Auto-learning/cognitive capabilities for the internal resource management can be easily incorporated.

• Support to the Cognitive Radio strategies thanks to the capacity to capture, in a coordinated way, relevant information from different layers of the radio and computing system.

PHAL-OE Concept & Layers

Previous figure tries to illustrate the PHAL-OE concept. On the bottom we can see the Hardware Layer where we can found several processors or Processing Elements (PE) physically interconnected among them. On top of the figure we can see the Abstract Application Layer where a graph of a radio application tries to define the required tasks (objects) and the data flow among them. In the Real Application Layer we can see the previous tasks but using the services or functionalities provided by PHAL-OE to build coherently the radio application. On th PHAL-OE Layer we observe that from the point of view of the tasks (objects) conforming the radio application all them only see the same platform, the PHAL-OE Platform, not being aware of any detail related with the hardware. On such layer, but from the hardware point of view, we can find the specific implementation of the PHAL-OE functions for each one of the processors used to build the entire Hardware Layer.