Jun 21 2011

Bullets From The Top Event on Cognitive Radio DySPAN 2011

The International Dynamic Spectrum Access Networks (DySPAN) symposium has emerged as the preeminent event to gather international economists, engineers, network architects, researchers and academic scholars together to share cutting edge research on and demonstrations of emerging wireless technology. DySPAN has since 2005 had a major influence on policy and technology research and development in the United States, Europe and Asia. The following bullets summarizes my key takeaways from DySPAN 2011:

  • Much focus was on TV White Spaces (TVWS). Studies on available TVWS channels in different regions was presented, and there seems to be a lot. The correctness of TVWS database estimations with different path loss models was presented by Ranveer Chandra from Microsoft by comparing with sensing results. None of the path models tested gave false positives, thought the accurateness of the path loss model is crucial for maximizing spectrum utilization. Furthermore, Victor Bahl from Microsoft also presented their TVWS network trials. Bahl stated that Microsoft and Google are enemies when it comes to search, but they are in bed together when it comes to TVWS because both want to provide connectivity to everyone to best extent possible. Overall, TVWS seems to be accepted by the DySPAN community.
  • Douglas Sicker from FCC presented the Spectrum Dashboard which also can be considered a tool to find available spectrum bands in the US. He also emphasized the importance of internationalization and encouraged a global market for dynamic spectrum access. He sends the message to DySPAN to work on this issue.
  • David Cleevely presented analysis of the value per MHz versus technology and concluded that unlicensed spectrum generates more value than licensed spectrum! Furthermore, he concludes that femtocells will revolutionize the mobile industry.
  • Krishan Sabnani from Bell Labs talked about spectrum virtualization and that a spectrum server is the final thing. He wants to use practices from virtualization in data computing. Base stations can be virtualized and run in the cloud.
  • Pierre De Vries presented a paper where he propose that operating rights should be articulated using transmission permissions and reception protections, defined probabilistically (the Three Ps). Such a policy would remove the zero tolerance of interference limit violation and ease the introduction of dynamic spectrum access.
  • Martin Weiss presented a paper where he focuses on the importance for the emergent regulatory policies to explicitly consider the requirements of secondary users, which not has been done yet because the secondary users not exists in meaningful number nor does they have any clear application. Furthermore he proposes to address this by explicitly consider the impact of the spatiotemporal properties of spectrum holes on the use decision by potential secondary users.
  • Mitola is worried about commercializing the white space databases since he doubts that we can trust the commercial sector?
  • In a panel about business perspectives on dynamic spectrum access, the urge of testbeds on white spaces was stressed by many. From an operators perspective Berna Sayrac from France Telecom mention that cognitive radio is much more than dynamic and opportunistic spectrum access, but that it can be used to manage our spectrum more efficiently. From a vendor perspective Lasse Wieweg from Ericsson mention that the most important for them is economy of scale, hence spectrum should be harmonized and he would like to see regulators worldwide to agree with each other and coordinate. He also states: do not be afraid of the higher bands such as 3.5 and 4.2 GHz.


Aug 4 2010

Key Experiences on Cognitive Radio from CROWNCOM 2010

The 5th international conference on Cognitive Radio Oriented Wireless Networks and Communications 2010 (CROWNCOM’10) was held in Cannes, France, from 09.06 to 11.06. Many experienced and respected researchers within the field of cognitive radio (CR) and the wireless system community attended the conference such as Joseph Mitola, Ian F. Akyildiz and Peter Olanders to mention some. Several hyper-relevant topics were discussed such as if CR is mature or immature, is CR ready for commercialization, what is happening within standardization, issues in physical resource sharing and what is the way ahead for CR after the first decade of CR research has passed.

Is cognitive Radio Mature or Immature?

Dr. Berna Sayrac from Orange Labs gave the first keynote about CR at orange labs where she at the end of her presentation stated that CR is an immature technology. Ian F. Akyildiz in the next keynote on the other hand, replied directly to this statement and stated that CR is a mature technology. This question was repeated and discussed throughout the conference and it seems that both statements are correct when viewed from different angles, thus Akyildiz may have correct in that cognitive radio is becoming more mature from a research perspective and Sayrac may have correct in that CR is immature from a commercial perspective as seen from the operator Orange’s perspective. Sayrac and Orange also see sensing as the biggest issue to be solved before CR can be commercially mature whereas Akyildiz among many technical issues especially emphasized that more research needs to be carried out on business models for CR.

Challenges in Commercialization of Cognitive Radio

This heading was the topic for a panel discussion where Yukitoshi Sanada (Keio University, Japan), Berna Sayrac (Orange, France), Jens Zander (KTH, Sweden), Peter Olanders (Ericsson, Sweden) and Joseph Mitola (Stevens Institute of Tech., USA) attended. Three prerequisites were listed for the commercialization of CR. First standards have to set the rules and protocols, second regulation has to allow CR operation and finally the business opportunities has to be in place. It was a common understanding that aspects of CR will first of all emerge as embedded functions in the various wireless standards that exist today and when standardization succeeds it will probably be seen in composite cognitive systems addressing cognitive radio and heterogeneous systems.

Olanders ended his presentation by plotting CR and SDR (software defined radio) in Gartners hype curve as illustrated by a reconstructed figure below,  which actually is similar to Nolan’s analysis in 2008. Mitola also agreed with Olanders view on the CR’s technology position between the peak and trough on the hype cycle and stated that this might be good since CR will move from the research era to the commercial era.CR and SDR on Gartners hype curve

Standardization of Cognitive Radio

Several times it was emphasized that CR is not synonymous with opportunistic spectrum access (OSA) or dynamic spectrum access (DSA) even though CR often is being associated with such technologies. CR is also important in heterogeneous networks where a mobile user can select between multiple wireless networks (WiMAX, LTE, DVB, WiFi, ..) and maintain multiple links simultaneously. Standardization of CR is not similar to other wireless systems such as WiMAX and LTE due to the fact that CR might be used within these systems in a heterogeneous manner as well as for secondary access to primary spectrum. Therefore standardization of CR should specify a CR framework instead of a CR system. The most relevant standards for CR mentioned were ETSI RRS, IEEE 802.22, IEEE 802.19.1, ECMA and SCC 41.

IEEE 802.22 has long been the most progressive standard for CR, but the rumor that the progress for this standard has slowed down lately was mentioned several times during the conference. Though Peter Olanders that has long experience within standardization also stated that “you never know, IEEE 802.22 might wake up again!”. The reason for no commercial IEEE 802.22 products or networks so far might be that the business case easily might become negative due to the services expected with distances up to 50 km in rural areas. Hopefully there will also be a focus on the low power portable or mobile applications within IEEE 802.22 in the near future. Another point is the white space database that needs to be in place for white spaces to be utilized in the US, but interestingly Google has announced that hey want to operate such a database. With the rumored speed of standardization in IEEE 802.11af, it will be interesting to see which of these will emerge in a commercial deployment first. However, IEEE 802.22 can be expected to be used by network operators, whereas IEEE 802.11af might be expected to be used by the general public in their homes as traditional WiFi and by network operators.

Lessons Learned

Many interesting subjects and ideas were presented in keynotes, panels and technical sessions at CROWNCOM 2010. Several technical issues were discussed and many discussions were concerned with commercialization of CR. A summary of some key lessons learned are:

  • CR seems to be moving from the research era to the commercial era.
  • Standardization of CR is more about specifying a CR framework instead of a CR system.
  • Three prerequisites were listed for the commercialization of CR: standards have to set the rules and protocols, regulation has to allow CR operation and finally the business opportunities has to be in place.