Understanding Power over Ethernet (PoE): Principles, Standards, and Implementation

Why PoE is Needed

With the proliferation of IP phones, wireless APs, cameras, and other networked devices, providing power via Ethernet simplifies installation, reduces cabling, and allows centralized power management, especially where outlets are scarce or devices are mounted in hard‑to‑reach locations.

How PoE Works

PoE systems consist of Power‑sourcing Equipment (PSE) such as PoE switches, and Powered Devices (PD) like APs or cameras. PSE detects PD presence, negotiates power class, and supplies up to 48 V DC over the cable.

Power Supply Modes

PSE devices are classified as MidSpan (external module) or Endpoint (integrated). They deliver power via Alternative A (data pairs 1/2 & 3/6) or Alternative B (spare pairs 4/5 & 7/8), but a single PSE must use only one mode while PD must support both.

Power Negotiation Process

Detection: PSE applies a low voltage to sense PD resistance.

Classification: PSE determines PD power class via resistance or LLDP.

Power Up: PSE ramps up to 48 V.

Normal Operation: PD draws power within PSE limits.

Shutdown: PSE monitors current and disables power on fault or disconnect.

LLDP‑Based Negotiation

IEEE 802.1ab defines a Power‑via‑MDI TLV that can convey power capabilities for IEEE 802.3af/at devices; Huawei adds a custom TLV to support IEEE 802.3bt parameters.

PoE Standards

IEEE 802.3af (2003) provides up to 15.4 W (12.95 W usable), IEEE 802.3at (2009) up to 25.5 W, and IEEE 802.3bt (2018) offers Type 3 up to 51 W and Type 4 up to 71.3 W, also supporting higher‑speed Ethernet.

Devices Supporting PoE

PoE‑capable equipment includes Huawei S‑series campus switches, wireless controllers, and AR routers; PoE capability is hardware‑defined and cannot be added via software upgrades.