docs: enhance README with detailed explanation of SIP and SDP architecture

README_cross.md

@@ -11,3 +11,68 @@ docker compose up -d --force-recreate
 # TODO
 
 - [ ] let user choose whether use mirror (use which mirror) when building Dockerfile 
+
+---
+
+Based on the logs and the **GB/T 28181-2022** standard you provided, here is the explanation:
+
+Yes, this is **SIP**, but the content *inside* the SIP message is **SDP (Session Description Protocol)**.
+
+While XML is used for *control* (like PTZ), **SDP** is used for **Media Negotiation** (setting up the video stream).
+
+### The Architecture from your logs
+1.  **SIP (The Envelope):** Starts the conversation ("I want to watch video").
+2.  **SDP (The Letter inside):** Describes technical details ("Send video to IP X, Port Y, using Format Z").
+3.  **RTP (The result):** After this SIP/SDP handshake finishes, the actual binary video stream (PS/H.264) starts flowing over a separate TCP/UDP connection.
+
+### Breakdown of your Log
+This log shows a **Real-time Live View** handshake.
+
+#### 1. The Request (SRS Server -> Camera)
+The Server asks the Camera to send video.
+
+```ini
+INVITE sip:34020000001320000001@3402000000 SIP/2.0
+Content-Type: application/sdp
+
+s=Play                         # "Play" = Real-time Live View (Standard 9.2.2.1)
+c=IN IP4 192.168.2.184         # The Media Server IP
+m=video 9000 TCP/RTP/AVP 96    # "Send video to my Port 9000 via TCP"
+a=recvonly                     # "I will only receive, not send"
+y=0911024252                   # **GB/T 28181 Special**: The SSRC (Stream ID)
+```
+
+#### 2. The Response (Camera -> SRS Server)
+The Camera agrees and tells the server its own details.
+
+```ini
+SIP/2.0 200 OK
+Content-Type: application/sdp
+
+c=IN IP4 192.168.2.64          # The Camera IP
+m=video 15060 TCP/RTP/AVP 96   # "I am sending from Port 15060"
+a=sendonly                     # "I will only send"
+a=setup:active                 # "I will initiate the TCP connection to you"
+y=0911024252                   # Matches the SSRC provided
+f=v/2/2560x1440/25/2/8192a/... # **GB/T 28181 Special**: Media Info
+```
+
+### Key Differences from Standard SDP
+GB/T 28181 modifies standard SDP with two specific fields mandatory for this protocol:
+
+1.  **`y=` (SSRC)**:
+    *   **Standard SDP:** Does not have a `y` line.
+    *   **GB/T 28181:** Uses `y` to define the **SSRC** (Synchronization Source). This 10-digit number is crucial because it marks every binary video packet sent later. If the binary stream headers don't match this `y` value, the stream is rejected.
+
+2.  **`f=` (Media Info)**:
+    *   **Standard SDP:** Does not have an `f` line.
+    *   **GB/T 28181:** Uses `f` to describe video parameters. In your log: `v/2/2560x1440/25...` means:
+        *   `v`: Video
+        *   `2`: Coding format (likely H.264 or H.265 mapped)
+        *   `2560x1440`: Resolution
+        *   `25`: Frame rate
+
+### Summary of Cooperation
+1.  **XML (SIP MESSAGE):** Used for "remote control" (PTZ, Query, Keepalive).
+2.  **SDP (SIP INVITE):** Used to *negotiate* the pipeline.
+3.  **Binary (RTP/PS):** The actual heavy video data that flows through the pipe created by the SDP.