Saturday, March 10, 2012
Hot Potato Routing and Real or Imagined Congestion
Several years ago when the Internet was beginning to grow, the matter of fair sharing of traffic loads arose. Because ISPs did not or could closely meter traffic as they can now, the possibly arose that some ISPs could hand off traffic to other carriers without detection and penalty. The term hot potato routing refers to the deliberate off loading of traffic onto another carrier’s network. The burdened carriers resented such “free riding” and emphasized that the hot potato routing carrier could not guarantee quality of service having handed off traffic to other carriers.
With the passage of time and the onset of real or claimed congestion, the hot potato resenting carriers have become hot potato routers themselves. Wireless carriers want to sell—make that give away—femtocells ostensibly so that subscribers can get better in-building signal penetration. This positive outcome occurs, because the carrier has had subscribers install mini-cellular radio towers on premises. But what kind of backhauling does the femtocell use? Some operate on cellular frequencies in effect retransmitting wireless signals. But others inject what would have been more wireless traffic into the Internet cloud via the wireless subscriber’s DSL or cable modem broadband connection. This is hot potato routing masquerading as signal enhancement. In reality the wireless carrier suffers less congestion and the need to cell split and install additional towers if it can offload traffic to other carriers.
Another example: Wireless carriers don’t mind—make that are happy when—subscribers substitute wi-fi minutes of use for cellular minutes of use. These very same carriers used to require equipment manufacturers to disable wi-fi access via cellular phones, so that subscribers had to use network minutes. With real or imagined congestion cellular carriers can reduce capital expenditures in more plant by offloading traffic onto another carrier’s network.
Bear in mind that such offloading often does not trigger a charge for the hot potato routing (traffic originating) carrier. While incumbent carriers, such a Verizon and AT&T, vigorously complain about free-riding VoIP operators, which find ways to inject traffic onto their networks without compensation, Verizon and AT&T do the same thing. Of course they can claim the need to do so results from the FCC’s inability to reallocate spectrum, or from unanticipated data service demand. What better way to shift the blame to the FCC or those pesky, gluttonous customers?
I appreciate that ongoing high capex harms carriers’ profits, but U.S. wireless carriers do not seem to be suffering at least in terms of Average Revenue per User (“ARPU”). These carriers have some of the highest ARPUs globally, and the elimination of unmetered and unlimited data plans promises even higher ARPUs going forward.
Skimping on capital expenditures accrues short term benefits, but it may foreclose longer term gains. If wireless carriers make the investment in 4G bandwidth and switching capacity, they can encourage subscribers to treat smartphones as the functional equivalent of wired computers. Additionally they can exploit technological and marketplace convergence that promote an IP-centric network serving an ever expanding aggregate demand for information, communications and entertainment (“ICE”) services. But the incumbent players dither: they threaten to become innovative and aggressive competitors, but back off. Cable operators realize the need to offer a wireless component in their bundle of services, but instead want to sell their spectrum to incumbent wireless carriers. Wireline incumbent carriers, such as Verizon, toy with the idea of offering video content, instead of serving as the conduit for the content managed and produced by others. But this week we hear that Verizon would rather partner with Netflix than compete.
Who wants to devote sleepless afternoons competing, investing and innovating?