Data vs. voice traffic


In the June 1999 issue, A. Michael Noll writes (in "Does data traffic
exceed voice traffic?")  that "claims of data and Internet superiority
in terms of network traffic are exaggerated and seem to be based
mostly in folklore."  He also argues that "voice traffic (measured in
bits) far exceeds data traffic and most likely will continue to do so
for the foreseeable future."  We published a paper on this subject,
"The size and growth rate of the Internet" (in the October 1998 issue
of the online journal First Monday, http://www.firstmonday.org).  The
statistics we work with are extensive, as are the details of our
arguments, since by necessity one ends up comparing apples and
oranges.  Therefore we will not attempt to summarize our work, except
to mention some of the conclusions.

We agree with Noll that, contrary to many claims, voice traffic still
exceeds data traffic.  We also agree with him that many reports about
astronomical growth rates of the Internet are exaggerated.  However,
whereas Noll estimates that "all forms of data communication on a
typical day in the U.S. gives ...  roughly 2% of the telephone
traffic," we calculate a substantially larger fraction.  The estimates
in our paper were that at the end of 1997, data traffic was between
15% and 25% of voice traffic.  Updating those estimates (using the
same methodology and data sources) to compensate for the growth
between December 1997 and June 1999, we find that today data traffic
is between 25% and 40% of voice traffic, and that data is likely to
overtake voice around the year 2002.

The main problem with Noll's estimates is that he did not collect any
measurements.  Rather, he used "back-of-the-envelope" calculations
based on asking his students about their activities to come up with an
estimate of 16.2 megabits of data traffic received per person per day
for his group at USC.  After reading his note, we did a slightly more
thorough study that took just a few minutes, yet was much more
comprehensive and precise.  An email message to colleagues at USC
revealed that statistics for USC's Internet traffic are available
online, at http:/foo.usc.edu/netstats.  These statistics, confirmed by
email correspondence with USC's network administrators, show that
during the school year, the average traffic from the Internet to USC
is about 20 megabits/sec.  Now USC has about 28,000 students, so
perhaps 35,000 people all told who use data communications.  Average
traffic of 20 megabits/sec, divided by 35,000 users, yields 49
megabits of data traffic received per person per day, or three times
as much as the average Noll computes for his group.

While Noll's estimates for data traffic appear low, his estimates of
voice traffic appear high.  He reports that on an average day, he
"participate[s] in well over an hour of long-distance calls."
However, FCC statistics, available at http://www.fcc.gov, show that on
average people in the U.S. (adults and children) participate in about
10 minutes of long-distance calls per day.  Although we do not have
any solid numbers for long-distance voice calls at USC, it seems safe
to assume, based on the FCC statistics, that they do not carry much
more traffic than USC's Internet connection.

Probably the main conclusion to be made from Noll's work is that he
and his students are atypical in making greater use of voice telephony
and less use of data communications than his colleagues.  Now USC is
not typical of the entire U.S. population, which is why our and Noll's
conclusion that data traffic for the nation as a whole still lags
behind voice traffic is valid.  However, data traffic is growing much
more rapidly, and even in the absence of real-time video on the
Internet it is likely to overtake voice traffic in a couple of years.

Kerry Coffman and Andrew Odlyzko 
AT&T Labs - Research