My Notes from chapter 17 of the CWNA study guide
History of PoE
- Eventually, some of the networking devices became small enough, both physically and electronically, that it became possible and practical not only to use the Ethernet cable to transmit data to the device, but also to send the electricity necessary to power the device
- With 10 Mbps and 100 Mbps Ethernet, two pairs are used for transmitting and receiving data and the other two pairs are unused. Gigabit Ethernet uses all four pairs of wires to transmit and receive data.
- When you are providing power to devices via the same Ethernet cable that provides the data, a single low-voltage Ethernet cable is all you need to install a networked PoE device
- Not only does this greatly reduce the cost of installing network devices, it also increases flexibility in terms of where these devices can be installed and mounted.
- Nonstandard PoE
- The initial PoE products were proprietary solutions created by individual companies that recognized the need for the technology
- The IEEE process to create a PoE standard began in 1999; however, it would take about four years before the standard became a reality.
- Proprietary PoE solutions often used different voltages, and mixing proprietary solutions could result in damaged equipment
- IEEE 802.3af
- PoE amendment to the 802.3 standard
- Defined how to provide PoE to 10BaseT (Ethernet), 100BaseT (Fast Ethernet), and 1000BaseT (Gigabit Ethernet) devices.
- IEEE Std 802.3-2005, Clause 33
- 802.3af amendment was one of four amendments that were incorporated into this revised standard
- IEEE 802.3at-2009
- The IEEE 802.3at amendment was ratifi ed in 2009. 802.3at is also known as PoE+ or PoE plus, since it extends the capabilities of PoE
- Two of the main objectives of the 802.3at Task Group were to be able to provide more power to powered devices and to maintain backward compatibility with Clause 33 devices
- As APs become faster and incorporate newer technologies, such as multiple input, multiple output (MIMO), they require more power to operate
- The IEEE 802.3at amendment is able to provide up to 30 watts of power using two pairs of wires in an Ethernet cable
- The 802.3at amendment defines PoE devices as either Type 1 or Type 2. Devices capable of supporting the higher power defined in the 802.3at amendment are defined as Type 2 devices, and devices not capable of supporting the higher power are defined as Type 1 devices
- IEEE Std 802.3-2012, Clause 33
- The IEEE revised the 802.3 standard again and created IEEE Std 802.3-2012
PoE devices (overview)
- The PoE standard defines two types of PoE devices: powered devices (PDs) and power-sourcing equipment (PSE). These devices communicate with each other and provide the PoE infrastructure.
- Powered device (PD)
- Either requests or draws power from the power-sourcing equipment.
- must be capable of accepting up to 57 volts from either the data lines or the unused pairs of the Ethernet cable
- must also be able to accept power with either polarity from the power supply in what is known as mode A or mode B
|Conductor||Mode A||Mode B|
|1||Positive voltage, negative voltage|
|2||Positive voltage, negative voltage|
|3||Negative voltage, positive voltage|
|4||Positive voltage, negative voltage|
|5||Positive voltage, negative voltage|
|6||Negative voltage, positive voltage|
|7||Negative voltage, positive voltage|
|8||Negative voltage, positive voltage|
- The PD must reply to the power-sourcing equipment with a detection signature and notify the power-sourcing equipment whether it is in a state in which it will accept power or will not accept power.
- If the device is determined not to be compliant, power to the device will be withheld
- If the device is in a state in which it will accept power, the PD can optionally provide a classification signature
- Type 2 devices perform a two-event Physical layer classification or Data-Link layer classification, which allows a Type 2 PD to identify whether it is connected to a Type 1 or a Type 2 PSE
- If mutual identification cannot be completed, the device can only operate as a Type 1 device
- If the device is not identified, the PSE does not know how much power the device needs; therefore, it allocates the maximum power
- If the device is classified, the PSE has to allocate only the amount of power needed by the PD, thus providing better power management.
- Link Layer Discovery Protocol (LLDP) is a standards-based layer 2 neighbour discovery protocol that can also be used for more detailed power classification.
- The maximum power draw of an 802.3af-compliant device is 12.95 watts, and the maximum power draw of an 802.3at-compliant device is 25.5 watts.
- Power-sourcing equipment (PSE)
- Provides power to the PD.
- Searches for powered devices by using a direct current (DC) detection signal
- After a PoE-compliant device is identified, the PSE will provide power to that device
- The amount of power provided by the PSE is greater than what is used by the PD
- This is because the PSE needs to account for the worst case scenario, in which there may be power loss due to the cables and connectors between
- The PSE and the PD.
- Once connected, the PSE continuously checks the connection status of the PD along with monitoring for other electrical conditions, such as short circuits
- When power is no longer required, the PSE will stop providing it
- Power-sourcing equipment is divided into two types of equipment endpoint and midspan.
- Endpoint PSE
- Provides power and Ethernet data signals from the same device
- Are typically PoE-enabled Ethernet switches
- The switches are typically access layer switches
- Some specialty devices, such as WLAN controllers may also function as endpoint PSE equipment
- Endpoint equipment can provide power using two methods referred to as
- Alternative A
- The PSE places power on the data pair
- Alternative A
- Alternative B
- Originally, Alternative B was designed to provide power on the spare unused pair of wires in a 10BaseT/100BaseTX cable,
- A 1000BaseT endpoint PSE can also use Alternative B to provide power to a PD by placing the power on two of the data 1000BaseT data pairs
- When 802.3af was initially ratified, 1000BaseT (Gigabit Ethernet) devices could receive PoE from only endpoint devices.
- Midspan PSE
- Acts as a pass-through device, adding power to an Ethernet segment.
- Enables you to provide PoE to existing networks without having to replace the existing Ethernet switches
- midspan PSE is placed between an Ethernet source (such as an Ethernet switch) and a PD
- acts as an Ethernet repeater while adding power to the Ethernet cable
- Originally with 802.3af, midspan devices were only capable of using Alternative B—and only with 10BaseT and 100BaseTX PDs
- With the ratification of 802.3at, midspan devices were able to use either Alternative A or Alternative B and they could provide support for 1000BaseT devices.
- Power-sourcing equipment pin assignments
- The PSE must have a medium dependent interface (MDI) to carry the current to the powered device (PD).
- MDI is essentially the technical term for the Ethernet cabling connector
- There are two valid four-wire pin connections used to provide PoE
- In each of these configurations, the two pairs of conductors carry the same nominal current in both magnitude and polarity
- Many devices are capable of automatically identifying and providing the crossover connection if needed
Planning and deploying PoE
- Power planning
- At maximum power for a PD, the PSE must be capable of providing 15.4 W or 30 W of power to each PoE device, depending on whether your devices require PoE+.
- This does not include the amount of power necessary for the switch to perform its networking duties
- A simple way of determining whether the power supply of the switch is powerful enough is to determine the size of the power supply for the equivalent non-PoE switch and add 15.4 watts for each PoE device that you will be connecting to the switch, or 30 watts for each PoE+ device you will be connecting to the switch.
- Careful planning is needed to ensure that enough power is available for all the PDs.
- When reading the power budget specifications of a switch, be sure to determine how many ports are PoE-capable.
- PoE ports can often also be configured with a priority level. Higher priority PoE ports take precedence for receiving power in the event that the PoE budget is exceeded
- Proper planning of the PoE budget to ensure that the budget is never exceeded is best practice.
- As the demand for PoE devices increases, the need to manage and troubleshoot PoE problems will also increase
- The more PoE devices that you add to the network, the more you concentrate the power requirements in the data centre or wiring closet
- As your power needs increase, electrical circuits supplying power to the PoE switches might have to be increased
- Even when there was an electrical failure, the telephone still worked and provided the ability to call someone
- This is a level of service that we have come to expect.
- As VoIP and VoWiFi telephones replace traditional telephone systems, it is important to still provide this same level of continuous service.
- To achieve this, you should make sure that all of your PoE PSE equipment is connected to uninterruptible power sources
- 802.11n or 802.11ac and PoE
- When 802.3at not available The most common method was to downgrade the MIMO capability of the 802.11n access points so that 802.3af power could be used
- The downside was that not all of the MIMO transmitter capabilities were being used by the APs.
- The good news is that almost all of the current generation of 3×3:3 dual-frequency 802.11n APs are indeed capable of running with full transmitter capabilities using 802.3af PoE.
- There is no reason you cannot use an available power outlet to provide electrical current to an AP. The downside is that most APs are deployed in areas where a power outlet is not conveniently accessible.
- The best way to power 802.11n and 802.11ac APs is to deploy a PoE+ (802.3at) PSE that is capable of providing 30 watts via an Ethernet cable