What is the most important intervention to be made to get cheaper, faster broadband to more people?
Firstly the brutal enforcement of the Facilities Leasing Regulations; Telkom should be facing crippling fines by this point in time.
Secondly ICASA through the processes has to introduce a wholesale per megabyte price for the “wireless local loop” which would be unbundled in this process. (The price runs from device to the point where the operator hands traffic over – so the operator is able by acting intelligently to avoid using its national backhaul capacity.)
The simple reality is that wireless based services in urban areas are exceptionally cost efficient as a last mile for moving across reasonably high volumes of data (per user), but in order for a wireless system to deliver high volumes of content you need a concentration of access points and ultimately the wired backbone of the network gets closer to the point of usage. Fundamentally it should be asked whether users should be using a wireless option in an urban area if they are transferring more than 100Gb in a month from a fixed point, inversely does ADSL (or other fixed line) orientated infrastructure makes sense for users who are transferring less than 3Gb in a month. In certain high density environments the mobility of most wireless offerings (3G) is the primary driver … Mobile operators have the power to introduce access at a reasonable rate by introducing products aimed for single tower usage (restricting its mobility and greatly adjusting planning processes). A fixed home broadband solution can be introduced by the evil two operators at R150 per month for 3Gb access with the operators pillaging in funds from the exercise.
Thirdly national policy needs to endorse and support open peering and the creation of network internet exchanges and other facilities around the country. The fact that Telkom provides handover for ADSL to ISPs only at three locations is ludicrous.
Fourthly the state should avoid putting money into individual operators or trying to put projects out for tender. Instead funding should be directed towards establishing an internet technologies financial services provider which provides suitable loans to the industry as a whole. [Whilst I haven’t fully researched the idea the principles applicable to co-operative banks (which the state has passed legislation for in 2007) might actually be beneficial in the industry – essentially communication licence holders (which are juristic persons) as well as natural persons working in the industry are eligible to be members of the co-op and applications for financing are received from licence holders. This approach moves it out of DoC possible problems – supervision for co-op banks lies with treasury and SARB and introduce further democratisation into the process]. If the financial service provider for internet development is truly open and transparent businesses which benefit from the network effects of broadband internet can readily be relied on to provide investments – it makes no sense for Google or the now listed Facebook to install a meter of copper on the last mile as they cannot charge for the access, but it makes sense for them to invest millions if it will see households having access to their service.
Fifthly construction and public work policies need to incorporate an open access electronic communications ethos. When SANRAL build a road (with or without corrupt tolls) they ultimately should build them such that fibre can be run without the road being dug up. Basic construction to includes ducts and so forth will avoid stupidity from municipalities seeking to halt fibre roll out.
Sixthly we desperately need Class-ECN licence holders to make a mark and build localized network infrastructure, every single local municipality should have one localized ECN operating. There is no point in Broadband Infraco building (or purporting to build) backbone infrastructure when nobody is going to be able to use it.
How do we deal with the problem of broadband services in rural areas?
A wireless last mile is generally presented as capable of ensuring rural area broadband with WiMax being particularly noted. The extent to which this is possible in the area protected for the SKA radio telescope project requires serious consideration. Whilst a major consequence of the SKA is (and will continue to be) the rolling out of fibre and connectivity for the project this needs to translate into broadband internet services for the communities which are in the area.
WiMax is well positioned for low population density areas (where other restrictions do not apply) but it is foolhardy to conduct planning applying the maximum distance possible and maximum speeds available.
More than the last mile (which can be many many miles) dealing with rural broadband requires a systematic construction of multiple network intensive nodes (data centres, POPs, “inx”s).
Whilst it may seem counter-intuitive sneakernets and IP by carrier pigeon should not be entirely scoffed at for moving high volume data which is not time sensitive in rural areas – largely to free up capacity for broadband internet services where latency matters. Most rural households have connections to towns or other nodes where wired internet connectivity at a high speed is possible. When considering offsite backup services for farming communities the benefits of the cloud need to be put into the perspective of whether solutions which sees the uploading of data by way of a continuous upload onto removable storage which is physically taken to a connection point for syncing with the server and where rural based businesses operate servers in datacentres (virtual or otherwise) and perform syncs using physical handover. In essence a farmer may have a torrent server (for legal downloads of course) which is on a high speed internet link away from the farm.