Time to Look Beyond TV White Space: Impression on CrownCom’13, July, Washington DC.

On July the 6th, I arrived at Washington DC. The atmosphere of the festival still lingered here: the stars & stripes hanging on every balcony; all trash bins full with picnic leftovers. This is the city that hosted the 8th Crowncom’13 and I was there to presented a paper on secondary system performance in TV White space.

Crowncom- International Conference on Cognitive Radio Oriented Wireless Networks-is a conference specifically focused on cognitive radio and spectrum sharing. It has not been known for its size. This year’s Crowncom is no exception, with only 38 papers out of more than 80 submissions are selected for presentation. but it hosts a few notable figures for keynotes and panel.

Perhaps the most celebrated speaker is the Prof. Anant Sahai from UC Berkeley. His speech centers around the question ‘Is wireless going to get a more modern regulatory layer soon?’ He argues that the legacy exclusive use of spectrum is moving towards a real time market, but the legacy system is not going away. In his opinion, the President’s Council of Advisors on Science and Technology (PCAST)  report is an intermediary solution to help the government keep up with its demand for spectrum. For instance, unmanned drones are being widely deployed by the army and national security agencies, and they require considerable more spectrum for wireless communication. Without the layered approach proposed by PCAST, the government’ budget simply can not support the market rate for spectrums. He also believes that licensed spectrum access has a promising prospect, because people would and had built networks without QoS guarantee and not all systems are like Wireless broadband that has insatiable appetite for spectrum. Furthermore, TVWS  is regarded as a success test run for the geolocation database method, although the aggregate interference problem still remains. Prof. Sahai’s view was later challenged in the panel discussion.

During the panel dissuasion, the BIG question was raise: what is left for cognitive radio research? More general topics are proposed as promising research directions, such as inter-system coexistence, sensing as a complement to database approach and a shift of focus from TV band to other ones like Radar band.

In fact these trends can already been seen in the current Crowncom publications. A quick comparison with last year’s crown com clearly illustrate how fast TV white spaces: lis losing its popularity: last year there are 43 oral presentation papers, out of which  12 are explicitly focused on TVWS: this year’s conference, in contrast, includes 38 papers, with only 5 papers focusing on TVWS.. The topics are now more diversified and the research interests on shifting towards more general subjects, such as resource management, energy efficiency and data mining in the context of cognitive radio.

One example on data mining is the presentation given by Prof Berna Sayrac from Orange Lab, who had given a keynote two years ago at crown com 10.  Her view is very much in line with the vision discussed by the CTO of noikia in his keynote in Dypsan12. As mobile devices full of sensors, operators can collect massive amount of data from user assiated measurement, far more than that from traditional road measurement, and, in fact, more than what the operator knows what to do with. therefore it is becoming ever more relevant to develop a strategy for the operator to handle the data efficiently and derive useful information.

I myself presented a paper ‘ on the capacity of wifi system in TV white space’, which was the offspring of the excellent thesis work by Yanpeng Yang. One of the key messages we hoped to deliver is that,  for a densely deployed  secondary wifi system, its capacity would be limited by secondary self interference and collision avoidance.

geo link rural

Contrary to the common belief/ fear that the the secondary system capacity would be limited by the aggregated interference caused to the primary in a dense deployment scenario, our finding instead points to the secondary self interference and the collision avoidance in systems with CSMA/CA type MAC as the real limiting factor. As one of attendee’s questions points out our high secondary user density assumption  is rather pessimistic, the secondary self interference would be less severe in a more general situations.  and as such it would achieve a higher capacity. On the other hand, the TV signal strength only affects the secondary capacity when the TV coverage is very poor and user density is very low.

In summary, Crowncom 13 hosts a smaller crowd than its predecessors, but it helps the attendee to better mingle with each other and provide the opportunity to avoid missing any single presentation.. The more diversified presentation topics is a clear signal for us to move on from TV WS back into the two decades long search for the ‘next big thing’ in cognate radio/spectrum sharing.

This entry was posted in Conferences. Bookmark the permalink.