---
title: Channel-wise Conditional Prompting (CCP)
url: https://www.emergentmind.com/topics/channel-wise-conditional-prompting-ccp
type: topic
---

# Channel-wise Conditional Prompting (CCP)

Channel-wise Conditional Prompting (CCP), as instantiated by the Channel State Prompt (CSP) mechanism in Prompt JSCC (PJSCC), is a differentiable architectural approach for integrating physical channel state information into deep learning-based joint source-channel coding (DeepJSCC). This design enables wireless semantic communication systems to adapt dynamically to varying signal-to-noise ratios (SNR) and channel distributions without the need for multiple specialized models, addressing fundamental generalization limitations in prior DeepJSCC architectures [2411.10178].

## 1. Architectural Structure of CSP in DeepJSCC

The CSP module is embedded within the PJSCC pipeline between Swin-Transformer stages in both encoder ($E_\phi$) and decoder ($D_\theta$) branches. In the encoder, an intermediate feature map $\mathbf{y}_1 \in \mathbb{R}^{c \times h \times w}$ from preceding Transformer Feature Extraction (TFE) blocks is passed to the CSP:

- **GAP+MLP Feature Extraction:** Global Average Pooling (GAP) reduces $\mathbf{y}_1$ spatially, and an MLP maps it to a compact vector $\mathbf{y}_2 \in \mathbb{R}^{c'}$ representing channel-aware features.
- **Prompt Retrieval:** Based on the current channel state $(\beta_{SNR}, \beta_{dis})$, a prompt tensor $\mathbf{p}_{ij}^e \in \mathbb{R}^{c' \times h' \times w'}$ is selected from the prompt set $\mathbf{C}_\gamma$.
- **Element-wise Modulation:** The compact feature $\mathbf{y}_2$ and prompt tensor $\mathbf{p}_{ij}^e$ are combined via element-wise multiplication to yield $\mathbf{m} = \mathbf{y}_2 \odot \mathbf{p}_{ij}^e$.
- **Spatial Expansion:** A convolution is applied to $\mathbf{m}$ to restore spatial dimensions.
- **Transformer Fusion:** This modulated feature is concatenated with the original $\mathbf{y}_1$, processed through a Transformer block and a convolution to yield the output feature $\mathbf{y}_3$.

Decoder-side CSPs are structurally identical, operating on feature maps within $D_\theta$ and using dedicated prompt tensors $\mathbf{p}_{ij}^d$. CSP modules are inserted between each pair of Transformer blocks along both encoding and decoding pathways [2411.10178].

## 2. Mathematical Formulation of Prompt Generation and Modulation

The CSP employs a discrete set of learnable prompt tensors $\mathbf{C}_\gamma$, indexed by all combinations of SNR levels and channel distributions. For $T$ CSP stages and $L$ channel states:

\[
\mathbf{C}_\gamma = \left\{\mathbf{p}_{ij}^e, \mathbf{p}_{ij}^d \mid i=1,\dots,L;\; j=1,\dots,T\right\}
\]

For input feature $\mathbf{y}_1$, channel-aware projection:

\[
\mathbf{y}_2 = \mathrm{MLP}\left(\mathrm{GAP}(\mathbf{y}_1)\right) \in \mathbb{R}^{c'}
\]

Retrieval and fusion use the matching prompt tensor $\mathbf{p}_{ij}^e$:

\[
\mathbf{m} = \mathbf{y}_2 \odot \mathbf{p}_{ij}^e
\]
\[
\mathbf{t} = \mathrm{Conv}(\mathbf{m})
\]
\[
\mathbf{y}_3 = \mathrm{Conv}\left(\mathrm{TransBlock}\left(\mathrm{Concat}[\mathbf{t}, \mathbf{y}_1]\right)\right)
\]

Only the prompt from $\mathbf{C}_\gamma$ corresponding to the sampled $(\beta_{SNR}, \beta_{dis})$ is used at each forward pass.

## 3. Feature and Prompt Fusion Strategy

CSP’s fusion scheme consists of three stages:

1. **Channel-wise Modulation:** Element-wise multiplication of the compacted feature vector and prompt tensor aligns semantic information with channel state.
2. **Spatial Merging:** A small convolutional operator reconstructs the modulated feature into full spatial dimensions for the next network stage.
3. **Transformer Fusion:** Concatenation with the original feature map is followed by a Transformer block and convolution for joint feature aggregation.

This design provides a fully differentiable, unified mechanism for channel-adaptive feature modulation, allowing the backbone network to learn to prioritize information transmission based on granular channel conditions.

## 4. Training Objective and Adaptability

The PJSCC framework with CSP modules is trained end-to-end using mean-squared error (MSE) loss between the input image $\mathbf{x}$ and its reconstruction $\tilde{\mathbf{x}}$:

\[
\mathcal{L}(\phi, \theta, \gamma) = \mathbb{E}_{\mathbf{x}, \beta_{SNR}, \beta_{dis}}\left[\|\mathbf{x} - \tilde{\mathbf{x}}\|_2^2\right]
\]

where $\tilde{\mathbf{x}} = D_\theta(\tilde{\mathbf{z}}, \mathbf{C}_\gamma(\beta_{SNR}, \beta_{dis}), R)$ and $R$ denotes the received signal realizations. Model parameters $\phi$, $\theta$, and prompts $\gamma$, are optimized jointly. Sampling from across SNR ∈ [1, 13] dB and both AWGN and Rayleigh channels during training ensures generalization to a diverse range of conditions without subsequent fine-tuning or separate model deployments [2411.10178].

## 5. Empirical Performance

Comprehensive evaluations demonstrate that the inclusion of CSP modules in PJSCC delivers superior image reconstruction quality over ablated and baseline systems. On CIFAR-10 under AWGN, CBR = 1/3, the reported PSNR (dB) at various SNR points is:

| Method                | 1 dB  | 4 dB  | 7 dB  | 10 dB | 13 dB |
|-----------------------|-------|-------|-------|-------|-------|
| w/ CSP (PJSCC)        | 30.30 | 33.28 | 35.92 | 38.17 | 39.97 |
| w/ AF                 | 29.64 | 32.64 | 34.84 | 36.83 | 38.58 |
| w/o CSP               | 29.56 | 32.50 | 34.68 | 36.65 | 38.40 |

Consistent improvements are observed on LPIPS metrics as well. Over full PSNR vs SNR curves and multiple datasets (CIFAR-10, Kodak, CLIC), PJSCC with CSP outperforms DeepJSCC, ADJSCC, WITT, and BPG+LDPC across both AWGN and Rayleigh channels. These gains indicate the efficacy of channel-aware prompting in enabling robust semantic transmission without a proliferation of models [2411.10178].

## 6. Memory Efficiency and Scalability

PJSCC with CSP modules achieves competitive inference times and computational cost. For high-resolution images, it requires 52.3 ms and 39.5 G FLOPs per sample; for low-resolution, 62.4 µs and 767 M FLOPs. This is comparable to WITT and an order of magnitude more efficient than ADJSCC and DeepJSCC ($>$10× FLOPs).

In terms of model storage, PJSCC-U can integrate both AWGN and Rayleigh channels in a single model, resulting in significantly reduced memory requirements compared to baselines that require separate models for each channel type:

| Model      | One Channel | Two Channels |
|------------|-------------|--------------|
| PJSCC-U    | 178 MB      | 216 MB       |
| WITT       | 120 MB      | 240 MB       |
| ADJSCC     | 131 MB      | 262 MB       |
| DeepJSCC   | 1469 MB     | 2938 MB      |

This storage compactness is attributed to the modularity and sharing of learnable prompts, eliminating the need for redundant parameterization per channel environment.

## 7. Ablation Studies and Robustness

Ablations establish the critical value of CSP prompts compared to both simple SNR concatenation (AF) and no prompting. Across all SNRs, using CSP increases PSNR by ≈1 dB over AF and ≈1.4 dB over no-prompt baselines. Each prompt tensor $\mathbf{p}_{ij}$ is sized to match the spatial and channel dimensions of the corresponding feature map stage (e.g., $c \times 32 \times 32$ in high-resolution blocks). Robustness tests confirm that the model generalizes seamlessly to previously unseen intermediate SNR values and both supported channel types, maintaining high reconstruction quality without additional fine-tuning or retraining [2411.10178].

---

CCP, as operationalized via CSP in PJSCC, constitutes a lightweight, fully differentiable prompt-based conditioning mechanism. It incorporates structured channel priors and achieves dynamic adaptation, memory efficiency, and high-fidelity semantic transmission across a broad spectrum of wireless conditions, all within a single unified DeepJSCC model [2411.10178].

Source: https://www.emergentmind.com/topics/channel-wise-conditional-prompting-ccp