An NBN retrospective - Part 4 ? X
Home GitHub Contact

Notes from 2022

This is Part 4, where we investigate just what the fuck is going on with the Fixed Wireless network. It's not pretty.

Part 1 focuses on the cultural deficiencies at NBNCo.

Part 2 focuses on the failings of FTTN.

Part 3 deals with the HFC network and ways in which it can be salvaged

Part 5 briefly touches on the economics of the NBN.

Part 6 hammers home the fact that all of this is a result of shameless and disgusting short sighted political manoeuvring by the Liberal Party to protect the interests of Rupert Murdoch and other corporate friends such as Telstra.

An NBN retrospective - Part 4.

It’s no great secret that the LTE component of the rollout has been spectacularly mismanaged. Utilising LTE to get high speed broadband access to properties too costly and too remote for wireline technologies was a good idea, however as with all things of this nature, cost cutting has led to a service so broken and so unreliable that many within its footprint elect to retain their ADSL services. Since switching to Fixed Wireless NBN, Whirlpool user Greg Lehey has experienced no less than 288 hours of downtime. Stories like this have emerged all over the country, with people reporting their services simply become unavailable during peak times. This is mostly due to bandwidth constrains imposed by the ineherent design of the NBN LTE network.

One of the biggest oversights in relation to the LTE network is the lack of backhaul capacity from each tower to its respective POI. As LTE is a shared medium, multiple users must share a single link back to the NBN Point of Interconnect, or POI. This is a colocation facility, where user traffic is handed off to the various ISPs’ networks. In order to cut costs, NBNCo elected to underprovision the amount of backhaul from each tower to the POI. Normally, it would be pretty trivial to increase backhaul capacity, as it would simply be a matter of increasing the bandwidth of the fibre optic link to the tower. However, NBNCo’s cost cutting method is so spectacularly stupid that this is simply not a possibility.

Instead of running a fibre optic cable to each tower for backhaul capacity, NBNCo has elected to use a wireless microwave link for many towers in a hub and spoke model. A microwave link is established between the POI and a master tower, which then divides the bandwidth of that link between multiple slave towers, which then connect actual end users to the network. This makes it incredibly hard to upgrade backhaul capacity to each tower, as it involves replacing expensive antennas or increasing the wireless bandwidth used, which often (always) isn’t possible due to extremely limited wireless spectrum available, and legally mandated power limits on wireless transmitters.

While NBNCo have claimed that there exists a list of towers and backhaul sites that require upgrading, there is little to no evidence that they have actually undertaken any such work. People are simply being left with the choice of either braving the peak hour congestion, or retaining their ADSL services and thus being forced to spend more than they need to. Regional areas within the LTE footprint are historically underserviced with DSL too, often only being able to get a connection from Telstra.

There are two solutions for remediating the LTE network. Obviously, the most pressing matter is getting sufficient backhaul to each tower in order to alleviate the peak hour congestion and inherent unreliability that comes with using a cascaded microwave link. The only way to do this is to run fibre optic backhaul to each tower. This is an absolute necessity if NBNCo ever hope to make Fixed Wireless a viable solution for regional broadband access, and should have been the only backhaul method specified in the Network Design Rules. Not only will it remove the weakest link in the network in terms of bandwidth capacity, it will also allow NBNCo to free up the wireless spectrum they currently use on backhaul and reallocate it to increasing bandwidth on the consumer side of the network.

Another pressing matter for the LTE rollout is tower oversubscription. Much like with HFC, there are simply too many end users competing for bandwidth from a single tower. Of course, with multiple towers sharing a microwave link back to the POI, this is simply a cascade effect of underprovisioning of backhaul capacity. However, assuming we have now supplied each tower with its own fibre optic link, we can address this issue by splitting down towers, again much in the same way an HFC node split works.

Assuming a 1:2 split, this would mean an effective doubling of available bandwidth for each end user between their premises and the tower. This increased bandwidth availability then allows NBNCo to offer faster plans across the network, making the same plans available across both wireline and fixed wireless technologies

Another minor consideration to make is the rising adoption of 5G, and the improvements it brings over LTE. At this stage, such an upgrade would be purely optional, and made in consideration of private competition. The expense necessary to cover regional Australia with 5G for either NBNCo or a private competitor would be ruinous, and so we will not consider this other than to remark that it is a possible (and likely) upgrade path in the medium to long term. With the bandwidth increases we have made above, LTE is still perfectly capable of delivering high speed broadband for some time to come, and any 5G upgrade would either be purely marketing or for increasing bandwidth efficiency to enable a greater number of customers to connect to the same tower.

Continue reading in Part 5.