This article is more than 1 year old
Britain's 4G is slower than Armenia's
Եկեք տեղափոխենք Երեւան:
The UK’s average 4G speeds are slower than Armenia’s, according to network performance monitor OpenSignal.
The firm’s international survey placed the UK in 39th spot internationally, four behind the mountainous Transcaucasian Republic, which can now boast an average download speed of 24.08Mbps. With an average 4G speed of 23.11Mbps, the UK has found itself two places behind Mexico.
Last year OpenSignal caused handbags to fly after it declared that the UK’s mobile broadband was slower than Peru’s. In fact Peru has excellent 4G coverage and 4G performance, but this drew criticism from rival network surveyors including the granddaddy of the market, GWS.* Peru has now fallen behind the UK in 4G performance.
Today OpenSignal published the results of its most recent UK survey, showing 4G coverage inching up to, or over 80 per cent for all four networks. That number is substantially larger than Ofcom’s coverage figure of 43 per cent – but it reflects actual user experience, OpenSignal pointed out.
Ofcom's figure referred to geographical coverage, and as such did not reflect actual user experience. “Our availability metric… accounts for how often users have access to an LTE signal no matter where they happen to be,” the firm noted. OpenSignal has used time connected to a 4G network as a proxy for coverage.
OpenSignal found O2 challenging EE’s leadership for coverage and performance in some regions: Wales and North of the Wash, but O2’s performance in London was pitiful, with the monitors recording an average throughput of just 8.8Mbps, compared to EE’s average of 27.4Mpbs. London was also O2’s slowest region: not a great advert for the Telefonica subsidiary.
3’s investment in the North East has paid off with the highest regional average, of 31.3Mpbs. Vodafone performed well on latency. Tutela reported this too back in December, in much more detail.
You can read the summary here.
Of the four companies that regularly report mobile network performance for the UK, two have relied on “crowd” methods, and two on “non crowd”, methods, which involved more rigorous but expensive surveying, resulting in fewer datapoint. Each approach has its merits. IHS Markit’s RootMetrics having started in the “crowd” camp and having mostly moved to using its own surveys. You can read the criticism and defence of the methods here – it’s really quite interesting. ®