Abstract
This paper describes the technical foundations for a future satellite-based digital communication network tailored for avionics. Key features include packetized transmission for multimedia efficiency, a power-efficient TDM downlink utilizing turbo-coded nonlinear modulation, and a CDMA uplink with adaptive interference suppression to mitigate the near-far problem. Additionally, we introduce multiple description image encoding at the transport layer to enhance end-to-end performance resilience.
Introduction
As aviation transitions toward data-intensive operations involving voice, telemetry, and high-resolution video, existing communication infrastructures face significant limitations. Satellite communication (Satcom) offers global coverage but is hindered by power constraints on the downlink, the near-far problem on the uplink due to large propagation delays, and the fragility of hierarchical data packets. This research proposes a hybrid architecture to address these challenges.
Downlink Architecture: Turbo-Coded CPM
To maximize power efficiency, satellite amplifiers must operate in the nonlinear saturation region. Traditional linear modulation schemes suffer significant degradation in this regime. Our design employs Continuous Phase Modulation (CPM), a constant-envelope signaling method that is inherently compatible with nonlinear amplification.
To achieve high reliability, we implement Bit Interleaved Coded Modulation (BICM) to pair CPM with powerful Turbo codes. This approach treats the complex M-ary CPM channel as a set of parallel binary channels, allowing the use of standard binary turbo decoding techniques.
Uplink Architecture: Adaptive CDMA
The aircraft-to-satellite link utilizes Direct Sequence Code Division Multiple Access (DS-CDMA). A primary challenge in this asynchronous environment is the 'near-far' problem, where signals from closer aircraft overwhelm those from more distant ones. Traditional closed-loop power control is ineffective due to the high round-trip latency of satellite links.
We propose a Differential Minimum Mean Squared Error (DMMSE) algorithm for blind adaptive interference suppression. This receiver-side technique allows the satellite to distinguish between users and cancel out interference mathematically, eliminating the need for rapid power control feedback loops.
Data Transport and Source Coding
Modern image compression (e.g., JPEG-2000) is often highly sensitive to packet loss. We propose a Multiple Description (MD) coding framework using Reed-Solomon erasure codes. By creating multiple 'descriptions' of equal importance, the system ensures that image quality scales with the number of packets received, rather than failing completely if a specific 'important' packet is lost.
Performance Evaluation
The proposed frameworks were validated through computer simulations using various signaling models, including CPFSK and 3RC CPM.
Key results include:
- Downlink Performance: Turbo-coded CPM (M=4) achieved a $10^{-5}$ Bit Error Rate (BER) with only a 1.6 dB SNR penalty compared to QPSK, a trade-off justified by the significant gains in amplifier power efficiency.
- Uplink Resilience: The DMMSE receiver maintained robust performance even when interferers were 20 dB stronger than the target signal. The system reached the 10% uncoded BER threshold (required for effective turbo decoding) at an $E_b/N_0$ of less than 5 dB.
Conclusion
By embracing the nonlinearities of satellite hardware and shifting interference management to the receiver, this hybrid CDMA/TDM architecture provides a robust path forward for avionics satcom. Future work will focus on refining frequency acquisition in high-Doppler environments and further optimizing joint source-channel coding strategies.
Summary updated