⚡ electrical engineering

Semantic Communications in 6G: Coexistence, Multiple Access, and Satellite Networks

This paper explores the integration of Semantic Communication with conventional Bit-based Communication in 6G networks, analyzing multiple access techniques for their coexistence, multi-modal frameworks, and specific applications in satellite networks to overcome bandwidth and channel constraints.

Original authors: Ishtiaque Ahmed, Yingzhuo Sun, Jingwen Fu, Alper Kose, Leila Musavian, Ming Xiao, Berna Ozbek

Published 2026-05-06

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: Ishtiaque Ahmed, Yingzhuo Sun, Jingwen Fu, Alper Kose, Leila Musavian, Ming Xiao, Berna Ozbek

Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

Imagine you are trying to send a message to a friend who is far away, but your phone has a very weak signal and you are only allowed to send a tiny, tiny amount of data.

The Old Way (Bit-based Communication):
Traditionally, we send messages by breaking them down into millions of tiny "bits" (like 1s and 0s). It's like trying to send a photo by mailing every single pixel individually, one by one. If the mail gets lost or damaged (bad signal), the picture comes out scrambled. To fix this, we usually need a strong signal or a lot of bandwidth (mailing capacity).

The New Way (Semantic Communication):
This paper introduces a smarter approach called Semantic Communication (SemCom). Instead of sending every single pixel, you send the meaning of the photo.

The Analogy: Imagine you want to tell your friend about a beautiful sunset. Instead of mailing a high-definition photo (which takes up a lot of space), you just send a short note: "Golden sky, orange clouds, calm ocean."
Why it helps: Your friend already knows what a sunset looks like. When they read your note, their brain "reconstructs" the image in their mind. You sent much less data, but the message got across perfectly. This is especially useful when the "mail" (the wireless channel) is bad or crowded.

What This Paper Explores

The authors are looking at how to mix this new "meaning-based" way of talking with the old "pixel-based" way, specifically for the future of 6G networks and satellites.

Here are the four main ideas they discuss, explained simply:

1. Mixing Old and New Users (The "NOMA" Party)
Imagine a crowded room where some people want to send detailed blueprints (BitCom users) and others just want to send the idea of a building (SemCom users). They have to share the same table (the same radio frequency).

The paper tests a technique called NOMA (Non-Orthogonal Multiple Access). Think of this as letting two people speak at the same time, but one speaks louder than the other.
The Finding: They found that if you let the "meaning" users speak when the signal is weak, and the "blueprint" users speak when the signal is strong, everyone gets their message through faster. It's like a smart traffic light that lets the right type of car go first depending on the road conditions.

2. The Satellite Challenge (The "Relay" System)
Satellites are far away, and the space between them and Earth is full of static and interference (like trying to talk to someone through a storm).

The Problem: For the "meaning" trick to work, the sender and receiver need to share a common dictionary (called a Knowledge Base). But in space, it's hard for a satellite and a ground station to have the exact same dictionary because they are so far apart and there are so many of them.
The Solution: The paper suggests using "Semantic Relay" satellites. Think of these as middlemen who act as a bridge. They help align the dictionaries between the sender and receiver so the "meaning" can be understood even if the connection is shaky. They also use AI to generate the missing parts of the picture if the signal gets lost.

3. Handling Many Types of Data (The "Multimedia" Smoothie)
So far, we talked about text. But what about sending a video, a voice recording, and a photo all at once?

The Challenge: Mixing these different types of data is like trying to blend a smoothie with ice, fruit, and soup. It's messy.
The Approach: The paper looks at frameworks that blend these different "flavors" into one unified "semantic smoothie." Instead of sending the video file, the audio file, and the text file separately, the system sends one compact message that says, "Here is a video of a dog barking." The receiver's AI then rebuilds the video, sound, and text from that single idea.

4. The Future (The "Smart" Network)
The paper concludes that while this "meaning-based" approach is amazing for saving space and working in bad conditions, it's not perfect yet.

The Catch: Sometimes you do need the exact original file (like a legal contract or a medical scan) where you can't afford for the receiver to "guess" the details.
The Verdict: The best future network will be a hybrid. It will use the "meaning" trick for things like streaming video or chatting (where a little guesswork is fine) and the "exact bit" trick for things that need 100% precision.

Summary

This paper argues that the future of 6G isn't just about sending more data faster; it's about sending smarter data. By teaching our networks to understand the "meaning" behind the bits, we can squeeze more information through crowded airwaves and keep satellites talking even when the connection is terrible. However, we need to be careful to mix this new smart way with the old reliable way so we don't lose important details.

Technical Summary: Semantic Communications in 6G: Coexistence, Multiple Access, and Satellite Networks

Problem Statement
The exponential growth of wireless subscribers and the resulting bandwidth constraints pose significant challenges for next-generation (6G) networks. Conventional Bit-based Communication (BitCom) systems are approaching the Shannon limit, particularly struggling with efficiency in low Signal-to-Noise Ratio (SNR) regimes. While Semantic Communication (SemCom)—which transmits extracted meaning rather than raw bits—offers enhanced Spectral Efficiency (SE) and intelligent resource allocation, it faces inherent limitations in signal reconstruction precision compared to BitCom. Furthermore, the integration of SemCom with existing BitCom infrastructures in heterogeneous networks, particularly within resource-constrained environments like satellite networks and multi-modal data scenarios, presents complex challenges regarding coexistence, resource allocation, and decoding strategies.

Methodology
The paper employs a multi-faceted approach combining theoretical analysis, mathematical modeling, and simulation to investigate SemCom integration:

Mathematical Modeling: The authors utilize a semantic rate model defined by $S = WI \frac{K}{L}\epsilon(K, \gamma)$ , where $\epsilon(K, \gamma)$ represents semantic similarity. Since a closed-form expression for $\epsilon$ is unavailable, the study adopts a generalized logistic approximation derived from DeepSC (Deep Learning-based Semantic Communication) outputs via data regression. This allows for tractable analysis of semantic rates under varying SNR conditions.
Coexistence Analysis (NOMA): The paper analyzes a two-user single-antenna uplink Non-Orthogonal Multiple Access (NOMA) scenario. In this setup, a "primary" user (near the Access Point) utilizes BitCom with constant power, while a "secondary" user (farther away) employs an opportunistic strategy, switching between BitCom and SemCom based on channel conditions. The study models Rayleigh fading channels and uses Successive Interference Cancellation (SIC) to evaluate ergodic semantic rates.
Satellite Network Framework: The authors propose a semantic relay architecture for satellite networks to address cascaded fading and Knowledge Base (KB) misalignment. This framework involves a JSCC encoder at the transmitter, a semantic relay satellite for signal enhancement, and a KB coordinator satellite to align the shared KBs between the transmitter and receiver.
Multi-Modal Survey: The paper surveys frameworks that integrate text, images, audio, and video into unified semantic representations using central fusion and task-oriented Joint Source-Channel Coding (JSCC).

Key Contributions
The paper outlines four primary contributions:

Satellite Integration: It explores SemCom for satellite networks using adaptive KB alignment and semantic relay architectures to mitigate bandwidth limits and harsh channel conditions.
Coexistence Performance: It provides a comparative performance analysis of SemCom and BitCom in a two-user uplink NOMA scheme. Specifically, it demonstrates how a secondary user can flexibly switch between paradigms to optimize resource usage while a primary user maintains constant BitCom transmission.
Multi-Modal Frameworks: It surveys multi-modal SemCom systems that fuse diverse data types (text, images, audio, video) via central fusion and task-oriented JSCC, highlighting both their potential and current challenges.
Future Directions: It identifies critical challenges and research directions for deploying semantic-aware systems in 6G, focusing on intelligent and energy-efficient network designs.

Results

NOMA Performance: Simulation results indicate that an "opportunistic" scheme, where the secondary user dynamically selects between BitCom and SemCom, consistently achieves the highest ergodic semantic rates compared to schemes using only one paradigm.
- At low to moderate bit-rate requirements ( $\bar{R} = 2$ bps/Hz), SemCom-only performance closely matches the opportunistic scheme, validating its efficacy in low-to-moderate SNR regimes.
- At higher bit-rate requirements ( $\bar{R} = 6$ bps/Hz), SemCom-only performance plateaus as transmit power ( $P_0$ ) increases, whereas the opportunistic scheme continues to adapt and outperform.
- BitCom-only schemes significantly underperform compared to SemCom and opportunistic approaches, particularly at lower transmit powers.
Decoding Strategies: The analysis suggests that in coexistence scenarios, decoding the bit-based signal first (followed by SIC for the semantic signal) is beneficial to prevent residual interference from affecting the bit user. Alternatively, leveraging shared KBs for semantic decoding can help reconstruct corrupted information.
Satellite Applications: The proposed semantic relay and KB coordinator architecture aims to overcome the difficulty of maintaining identical KBs across wide satellite coverage areas and to mitigate severe cascaded fading.

Significance and Claims
The paper positions SemCom as a transformative paradigm for 6G that moves beyond the limitations of Shannon's classic information theory by focusing on semantic accuracy and task-oriented transmission.

Complementarity: The authors argue that SemCom is not a replacement for BitCom but a complementary technology. BitCom remains superior for exact signal reconstruction at high SNR, while SemCom excels in low SNR and bandwidth-constrained environments.
Heterogeneous Networks: The integration of both paradigms via advanced multiple access techniques (like NOMA) is presented as a key enabler for next-generation networks, allowing for flexible resource allocation among diverse user types.
Satellite Resilience: In the context of Non-Terrestrial Networks (NTNs), the paper claims that SemCom, particularly when combined with AI-driven semantic extraction and generative models, can significantly reduce transmission latency and bandwidth overhead while maintaining data integrity for applications like Earth observation and disaster relief.
Future Outlook: The work concludes that realizing the full potential of SemCom requires further research into standardized semantic metrics, the optimization of semantic KBs, and the development of energy-efficient implementations suitable for resource-constrained devices.

What This Paper Explores

Summary

More like this