The instinct, when a cell is congested, is to add radio: more spectrum, more sectors, a small cell on the corner. It is the part of the network an RF plan makes visible, and the part a vendor is happy to sell. In most East African densification cases, it is also not the binding constraint.
Capacity is a transport problem wearing a radio costume
A modern mid-band 5G radio can deliver a great deal of capacity into a sector. Whether a user ever sees it depends on what sits behind the radio. A macro site carrying real load needs backhaul in the multi-gigabit range; a dense cluster of small cells multiplies that requirement across every lamp-post you add. The radio is the cheap, fast part. The transport to feed it is the slow, expensive part.
That is why adding radios to a congested area often moves the bottleneck rather than removing it. If the aggregation ring behind those sites is already full, a new small cell inherits the same ceiling — now shared among more radios. The coverage map turns green; the throughput a user actually sees does not move.
The three constraints that decide a densification plan
Backhaul. A small cell needs fibre or a high-capacity microwave hop, and in most of the region fibre thins out quickly outside the metros and the trunk routes. Microwave buys you reach, but it needs line-of-sight, spectrum, and its own aggregation — and it caps out well below what a fully loaded 5G sector can produce. The honest question for any densification plan is not “where do we need coverage” but “what feeds each new site, and what does that path cost.”
Power. A site is only as available as its power. Where grid supply is intermittent, every site carries backup, and backup is capex on day one and opex every month after. Densification multiplies the number of small, distributed power problems you are now operating.
Site economics. Rents, wayleaves, access, and permitting set the real per-site cost, and they do not fall just because the radio is small. A hundred micro-sites is a hundred negotiations, not one.
What we look at
When we review a densification or deployment plan, the radio plan is where we start, not where we stop. We test the link budget and interference assumptions for realism, then we go to the part the plan usually under-specifies: the backhaul per site, the power design, and the total cost per delivered gigabyte rather than per site built. The distance between advertised throughput and delivered throughput is almost always a transport-and-power story, and it is visible early if you look for it.
The radio is rarely where a rollout dies. The link feeding it usually is — and the weakest link, not the strongest radio, sets what the network delivers.