IN-ORBIT TEST BED FOR AI USING HARDWARE-ACCELERATION-ASSISTED SIGNAL PROCESSING OF RANDOM-ACCESS PROTOCOL SIGNALS (ARTES AT 3E.008)

Description

The objective is to design, build and test an in-orbit test bed, comprising an RF receiver/transmitter, Artificial Intelligence (AI) hardware and associated signal processing algorithms, for the reception of low power random access protocol signals using artificial intelligence and machine learning algorithms that can either be trained on the ground or self-trained in orbit. The in-orbit testwill apply and refine these algorithms in the real-time, dynamic signal environment produced by the ground elements and interference sources. Targeted Improvements: Twofold increase in the number of random-access protocol signals processed compared to the currentstate of the artDescription:Post processing of modern low power random access protocols (such as ADS-B) have been shown to improve the number of messages recovered. Artificial Intelligence (AI) and Machine Learning (ML) algorithms are currently being developedto support these post processing techniques for several similar applications. Improving the success ratio of recovering messages reduces the number of satellites required in a constellation and improves the viability of telecommunication business cases targeting this market. This activity will develop, launch and operate an in-orbit test bed which can be used to test the performance of AI/MLalgorithms for post processing of low power random access protocol signals generated by existing ground devices. The in-orbit test bed will be deployed in an operational environment that includes the presence of a large number of transmitters, multi-path and propagation effects, interference sources, real antenna characteristics and varied terminal locations.The in-orbit test bed will select and embark a leading-edge AI hardware accelerator to process and react to received signals from established random access protocol ground devices. It will be designed to support the demonstration in-orbit of AI algorithms, such as, but not limited to, those being developed under on-going ARTES AT activities.Specifically, the in-orbit test bed will be developed to: Support the reception and recovery of messages from at least one established family of random-access protocol devices. Include the ability to transmit signals to ground devices to test protocols which can respond to the changing environment (variable modulation and coding, re-transmissionrequests, or frequency change requests). Support post processing with and without AI/ML techniques, allowing the performance improvement offered by these algorithms to be quantified. The ability to support their uplink to and execution on the spacecraft is alsorequired. Procurement Policy: C(1) = Activity restricted to non-prime contractors (incl. SMEs). For additional information please go to: http://www.esa.int/About_Us/Business_with_ESA/Small_and_Medium_Sized_En…

Tender Specifics