Enterprise Deployment – Ethernet and the Physical Layer

April 11, 2016

ethernetappleTo begin the discussion on enterprise transport, we’ll start at the bottom with the humble cabling that transports data traffic and work our way up. Please refer to the start of our series here if you have not read it.

In most office settings, Category 5e unshielded twisted pair “UTP” cabling is the dominant physical layer that carries Ethernet connections from Ethernet Switches to attached computers or Wi-Fi Access Points. It has the necessary bandwidth to support a 1Gbps full-duplex Ethernet connection (IEEE 802.3 1000BaseT) while supplying DC power from the Ethernet switch for Power over Ethernet plus (PoE+ IEEE 802.3at) attached devices like the E-RAN’s Radio Nodes.

Category 6a and 7 cable plants are unusual in most enterprises as their costs are much higher (50% and 100% higher, respectively) with no perceived benefits. Why? The majority of computers and access points attach at no more than 1Gbps and, where a 10Gbps attachment is needed, it usually is an optical connection choice due to copper transport constraints that limit distances.

Now, let’s connect this to indoor small cell architectures. For those who are not familiar with the E-RAN architecture, please refer to this short video before proceeding forward.

The E-RAN has a C-RAN split that implements, via a SoC, Layer 1 & 2 in the Radio Nodes (shifting analog processing to the edge of the network), and a standard PoE+ 1Gbps Ethernet for IP/IPSec fronthaul connectivity. This design choice means that the data traffic over the fronthaul network is limited to:

  • Encrypted GTP flows between UE’s and the mobile core.
  • E-RAN housekeeping between the Services and Radio Nodes.

It also allows the E-RAN to use whatever the available data cable plant that the enterprise has available in the site.

For alternate indoor architectures that have originated by downsizing outdoor macro-cellular technology, they require a more expensive Category 6a or 7 shielded twisted pair cable type. The cabling is used as a wideband transmission line to replace coaxial cable or fiberoptic cable between the box in the Telecommunications Room and a radiating element. It’s important to note that this is not Ethernet, and has no relationship with the enterprise infrastructure.

Additionally, if the enterprise and proposed indoor system use different cable plant (enterprise uses Category 5e and indoor system uses Category 6a), it will require the establishment of an additional patch panel in the Telecommunications Room. This is a notable issue due to rack space constraints in most enterprise Telecommunications Rooms as most of them are jam packed with their own enterprise transport infrastructure.

At a high level, this is the comparison:

telecomcloset

In following the theme of re-opening the discussion on how indoor cellular should be built, infrastructure that is common to all wireless is key to future enterprise use of small cells. Enterprises will not want to acquire solutions that require construction of a parallel network.

E-RAN is “enterprise IT friendly” by using:

  • Incumbent cable plant.
  • PoE+ Ethernet for Radio Node power and fronthaul.
  • IP/IPSec for fronthaul. Load is UE traffic plus a 10% overhead for E-RAN housekeeping.

In the next installment, we will delve into the dimensions of Telecommunications Room logistics and constraints.

– Art King, SpiderCloud Wireless, Director of Enterprise Services & Technologies

Twitter: @ArtKingg
Visit our Enterprise IT site @ http://SpiderCloud.com/EInsider

Other posts from the Enterprise Deployment Series:

  1. Enterprise Deployment – Setting the Stage
  2. Enterprise Deployment – Ethernet and the Physical Layer
  3. Enterprise Deployment – Telecommunications Room
  4. Enterprise Deployment – Telecommunications Room Interconnection
  5. Enterprise Deployment – Equipment Room
  6. Enterprise Deployment – Backhaul
  7. Enterprise Deployment – Summary

Leave a Reply

Your email address will not be published. Required fields are marked *