[kictanet] IG Discussions- Day 2 of 10: Infrastructure Issues-Submarine Cables

[Michuki Mwangi]

Hi Walu, et al,

Access:

I believe that we need to move beyond the marketing hype into the
realities. IMHO there's no broadband in Kenya. In my basic understanding
Broadband means affordable and fast connectivity. What we have is
average-to-unreliable connectivity that costs more than its worth.

I believe the ongoing investments into wireless solutions while they
work will continue to hold us into the pre-broadband phase for a long
time to come.

Mobile broadband seems to be a working solution but their ability to
scale will continue to pose challenges on quality.

Therefore investment in fixed infrastructure solutions like FTTH and
FTTC are going to be important considerations to be made if we are to
leap into broadband phase.

At this point, it would be ideal if regulators would open up the last
mile (within residential spaces) for anyone interested to invest in that
space with fixed infrastructure.

The de facto point here is that for a long time residential access has
been neglected for a long time. The focus has been mainly at CBDs until
the mobile 2G and 3G solutions were introduced.

Maybe policy has a role to play in providing incentives for players who
make efforts in providing fixed infrastructure in residential locations.

Affordability:

The pricing will still remain high until we have a inexpensive local
loop solution.

As far as the cables go it will be good to know more details before
folks can make conclusions. For instance is the 500-100USD per month for
bits or bytes?. It would be good to know what pricing model the cable
operators want to have in place for instance does one buy a circuit from
point A to B or does one buy a circuit with Internet on it already?.

Selling circuits are the ideal way to go hence it allows operators and
others to buy large pipes STM1 (144 Mbps) from say Nairobi to London.
Its cheaper to buy transit capacity from London which goes for about $10
Per MB per month the other costs will be for the circuit back to Nairobi
and that can be low if talking of big pipes. But then the models are not
clear to me so i cannot give conclusions on what is.

Content:

The only way we can drive up content availability is by placing the
requisite infrastructure. As it stands, few host content locally the
pricing is beyond and facilities are few hence higher pricing.

In Kenya my estimated total collocation/hosting capacity available is
well under 2,000m2 and am being kind here. In Europe one company
Telehouse has 30,000m2 of datacenter space 20,000m2 of which is in the UK.

We have afew factors in favor of Kenya for instance cheap land and
professional Labor (we cannot compare to EU rates). The only expense is
power but we have abundant solar energy that can serve the demand for
green data-centers one would wish to build. As some of you may know
Europe ran out of hosting space back in 2006 and has lead to the
emergence of Asia as the next collo destination.

Its upon the region to realise that we could be home to the future
worlds data centers with low cost energy, land and labor in our favor.

That way we have both local and international content hosted locally and
this will drive costs of access to lows never imagined.

Quality:

Quality of service will improve;

1) With better and affordable local loop infrastructure (not over
subscribed wireless base stations) - i understand that the regulatory
licensing structure had a negative impact on this as operators had to
pay an annual fee per installed site?. Probably fixed wired solutions
that have no recurrent costs are the way to go?

2) Increased percentage of locally hosted/accessed content (peering
traffic) which is subject to less contention and better access speeds

3) better pricing models on transit capacity which will reduce the
contention ratios given by service providers

4) More training of engineers in building scalable services and routing
infrastructure.

Regards,

Michuki.