[Click here to see part 1 of this 2-part article]

“There are better ways of assuring very high reliability and very high quality of service than selling spectrum,” said Rajeev Krishnamoorthy. He was referring to the idea that allowing radios to self-organize will provide for a more efficient use of spectrum than through existing allocation methods. Devabhaktuni Srikrishna points out that their modeling suggests that this approach to spectrum allocation could double the throughput to existing networks.

Krishnamoorthy and Srikrishna also discuss possible pilots projects.

Click here to see part 1 of the above video interview.

FirstNet Would Make a Good First App

One application that policy makers should look at for this approach to would be the FirstNet network. Equipping each of the approximately 1.5 to 2 million first responders (police officer, paramedics, firefighters and EMTs) in the U.S., with a $500 smart phone (not too unreasonable, considering very capable android phones can be had for less $250)  outfitted as a node in a SocialMesh network would only cost $1 B, as compared to the $7 B that has been allocated and the $10 B that some suggest is necessary for FirstNet. This alone would eliminate the need for auction of broadcast spectrum in order to fund FirstNet.

In  follow up email, Srikrishna reinforced the notion that FirstNet is a natural approach for a SocialMesh approach.

“It would make much more sense for FirstNet to pursue a SocialMesh approach based on frequency-agile smart phones rather than waste decades trying to build and standardize a cellular infrastructure that piggybacks on carrier networks. Public safety and police have been asking for reliable, high-speed P2P/D2D communications for at least a decade, and we (collectively as an industry) have not delivered on that need — why? Right now, public safety agencies are fighting a standards battle to get D2D communications embedded into the future versions of the LTE standards, but since it’s not a priority for carriers, who knows if that will be implemented in commercial smart phone chips even if it makes its way into the LTE standards.”

He goes on to point out that the power of the SocialMesh approach is the crowd.

“The only reason FirstNet needs so many base stations in the first place is that there are not always enough first responders around to relay the packets back to the Internet (1-2 million total in the US) in a time of need. What if hundreds of millions of citizen smart phones that are already roaming around could act like FirstNet base stations? We would need a lot fewer fixed base stations to begin with.”

Long term, I think it makes more sense for the FirstNet vision to expand to include all citizens

1. They can add many more nodes to the Crowd Cloud than public safety officers or vehicles can alone.
2. Often times the need during disasters is to enable communication to/from citizens for weeks and months after the initial event. For example this is what happened during Sandy and Katrina.

In this spirit, some people including our former FCC chairman Genakowski have aptly proposed we should include citizen smart phones using the Wi-Fi radio already in the smart phone,

http://www.technologyreview.com/view/515781/former-fcc-chairman-lets-test-an-emergency-ad-hoc-network-in-boston/

I recommend this well-written article and I fully agree with the thrust and purpose of his proposal. However my only comment is that using 20 dBm smart phone Wi-Fi interfaces for this purpose represents a marginal benefit. This is due to limited range of the Wi-Fi in smartphones (100-200 feet in practice). Speaking from deep experience working with outdoor Wi-Fi, the main reason Tropos routers have 36dBm EIRP and are mounted high up on poles is that it takes a much larger link budget to power through urban jungles.

It is too optimistic to expect that a crowd cloud based on smart phone Wi-Fi will be there for us when public safety first responders or citizens really need it during disasters or crises. As our paper on the Mobile Crowd Cloud shows, to fully leverage the spatial diversity present in the Crowd Cloud and create a usable disaster-proof network for everyone, a more capable radio interface (an LTE-like radio or better) is needed.”

[Added 10/14/16 – goTenna’s latest offering (referenced at the bottom of this article) appears to feature the attributes of the social mesh network outlined above and could be an interesting approach for a FirstNet-style network.]