从 Modbus 到 MQTT：协议栈的降维打击

在工业自动化领域，Modbus 已经服役了数十年。这种基于寄存器读写的串行协议曾经是连接现场传感器的黄金标准。但在智能网络时代，继续死守 Modbus 正在阻碍工业系统走向智能化。

将 MQTT 引入以太网边缘端点，将带来一场协议栈维度的“降维打击”。

---

1. 为什么 Modbus 在云端计算时代落伍了？

1. 主从轮询 (Poll) 机制的低效：Modbus 要求主站（网关/PLC）必须不停地、周期性地去“询问”从节点是否有数据变化。如果下挂 100 个节点，总线大部分带宽都会被无意义的问答消耗。
2. 缺乏数据自解释性：Modbus 只传输裸露的 16 位寄存器数值。如果没有专门的点表对应，你根本不知道寄存器 40001 里装的到底是温度、湿度，还是阀门状态。
3. 安全赤字：Modbus 没有任何认证和加密，任何人都可以随意对 PLC 发送写入指令，引发产线灾难。

---

2. MQTT + TLS 1.2 在边缘感知的降维打击

KeeTion 模组在 10BASE-T1S 以太网层之上，直接运行了原生的 MQTT + TLS 1.2 客户端。

• 事件驱动发布/订阅 (Publish/Subscribe)：节点只有在传感器数值发生改变，或发生超限警报时才会向总线“发布（Publish）”消息，其余时间处于安静监听状态。极大节省了网络带宽。
• 自解释 JSON 载荷 (Payload)：直接发送符合云端物联网平台规范的 JSON 文本（如 {"temperature": 23.5}），不需要任何协议适配和翻译。
• 工业级的安全传输：借助微型化的 mbedTLS 密码引擎，从传感器端点即可进行双向证书校验，杜绝数据伪造与非授权控制。

From Modbus to MQTT: A Protocol Stack Paradigm Shift

For decades, Modbus has been the workhorse of industrial automation. This serial protocol based on register read/write was once the gold standard for connecting sensors. However, in the era of smart networking, sticking to Modbus is hindering the evolution towards smarter industrial systems.

Bringing MQTT to the Ethernet edge brings a generational paradigm shift to the protocol stack.

---

1. Why is Modbus Obsolete in the Cloud Computing Era?

1. Inefficient Master-Slave Polling: Modbus requires the Master (PLC/Gateway) to continuously poll slave nodes. If 100 nodes are on the bus, most bandwidth is wasted on repetitive questions and empty answers.
2. No Data Self-Description: Modbus only transfers raw 16-bit register values. Without an explicit point list mapping, it's impossible to know if register 40001 contains temperature, humidity, or valve states.
3. Security Deficit: Modbus lacks any authentication or encryption. Anyone on the bus can issue malicious write commands, leading to plant disasters.

---

2. MQTT + TLS 1.2: A Higher Dimension of Edge Networking

KeeTion's module runs a native MQTT + TLS 1.2 client directly on top of the 10BASE-T1S Ethernet layer.

• Event-Driven Publish/Subscribe: Nodes only publish messages when sensor data changes or anomalies occur, remaining silent otherwise. This saves bus bandwidth.
• Self-Descriptive JSON Payload: Directly sends structured JSON (e.g., {"temperature": 23.5}) matching cloud specifications, eliminating custom translators.
• Industrial-Grade Security: Utilizing a lightweight mbedTLS engine, the node performs mutual TLS certificate authentication directly at the sensor level, ensuring data integrity.