Leverage Geotagged Social Data for Public-Safety LTE
By Robert Dew
Tuesday, May 26, 2015 | Comments

Commercial wireless carriers, utilities, public safety and other organizations are migrating to Long Term Evolution (LTE) for wireless broadband data. LTE allows for interoperability, a flat IP architecture with higher data throughput and lower latency, wider breadth of real-time applications and services, more sophisticated quality of service (QoS) and priority mechanisms, and economies of scale.

However, optimal placement of LTE evolved Node B (eNodeB) sites to ensure the best coverage, capacity and user experience to meet network design availability and reliability requirements and budgetary constraints is a challenge. Macro cell site placement considerations can include the selection of existing site assets for LTE equipment installs, which LTE sites will require more LTE equipment for capacity augment, and where and when new sites need to be constructed to fill in coverage or capacity holes based on minimum cell edge uplink and downlink data throughput requirements.

For small cells, which provide coverage (typically between 10 – 200 meters) and capacity while mounted on structures such as light poles or operating as residential eNodeBs, pinpointing exact geographic locations where they are required is a must. Small cell deployments augment capacity of existing LTE macro cell networks both outdoors and indoors. Commercial wireless deployments could more than triple the number of macro and small cell sites in these heterogeneous networks (HetNets) because of capacity requirements and smartphone growth. That would result in an exponential growth in the complexity of site planning. In that instance, maximizing the value of these resources would require automated planning with the highest accuracy possible down to centimeters, while using 3D vector data.

The Value of Geotagged Social Data
Companies such as Google and Amazon rely on big data analysis to develop, target and enhance services for their customers. For the past three to four years, savvy commercial LTE design engineers have been using design tools that collect, aggregate and leverage geotagged social data to create highly accurate traffic heat maps. These maps determine optimal placement of LTE macro and small cells to provide capacity where it is needed based on where public users move and cluster in a given 24-hour period.

For example, a small cell LTE design in a portion of Grant Park in Chicago strategically placed low-power small cells exactly where they would be of most use — where the public is most likely to be based on a population density weighted with geotagged social data. The placement of small cells depends on other design and deployment criteria, such as backhaul and power availability, physical and network security, network timing, structural integrity, interoperability, and in-band and out-of-band frequency interference.

Current population estimates in most LTE designs are based on census grids, which assume where the public statically lives and works. However, these census grids and blocks do not account for where the public moves or clusters during the day, which can underestimate how many people are most likely to be at a certain location at any time during a 24-hour period. Specifically, within census grids and polygons, uniform population density is assumed. Therefore, there is no indication of how the public clusters within the polygon or grid.

To derive a highly accurate traffic map, an LTE design must be granular enough to account for localized concentrations of data usage, critical for macrocell and small cell planning. Clusters of high data usage at the cell edge will invalidate the assumed average cell spectral efficiency (CSE), and therefore, the assumed capacity of each macrocell.

Leveraging geotagged social data collected during an extensive period of time along with super granular mapping data from light detection and ranging (LiDAR) with bin sizes at 1 meter down to 10 centimeters that can pinpoint usage will provide a more accurate estimate on where people are during the course of the day. The actual spatial distribution of the public must be taken into account based on social data showing real user behavior. Major sources of mobile geotagged social media data are Twitter, Facebook, FourSquare, Flickr and Google Plus.

New York City illustrates the importance of understanding where people actually are vs. their census coarse distribution. The dynamic population of Manhattan is different from the census-data-based population of Manhattan. With a census-measured residential population of 1.6 million, Manhattan more than doubles to a daytime population of 3.9 million. Peak population events, such as the Macy’s Thanksgiving Day Parade, often push Manhattan’s daytime population to 5 million, depending on the conditions and circumstances.

The same logic holds true for many other areas such as Disney World or areas with temporary events, such as Burning Man in Nevada.

Public-Safety Benefits
Geotagged mobile social data can help public safety in various ways. Social data ensures overall wireless broadband coverage objectives include where the public lives, moves, clusters and is most likely to be during the course of the day. It also can help prioritize placement of macrocells now and small cell capacity in-fill during initial and follow-on deployments, leveraging user equipment (UE)-based metrics software once the network is operational.

Detailed population data also can help prioritize in-building objectives as the network begins to infill indoor coverage and capacity. Further, public-safety planners can determine where hardening of sites should be prioritized based on areas where there could be the largest loss of life because of manmade or natural disasters. These areas should be prioritized via geographic information systems (GIS) data in the LTE coverage objectives, as well as ultimate design leveraging geotagged social data along with the latest applicable census data, CAD 9-1-1 incident data, geographic hazards, critical infrastructure, and other public-safety areas of operations. Leveraging geotagged social data will ensure eNodeBs are placed in areas where the highest probability of loss of life may occur and is more accurate than using census data alone.

In a temporary event, such as Burning Man, a global composite total population density, combining and weighting census population data with mobile social data down to a 50-centimeter resolution, shows users in a coverage objective area. As a result, there is no need for public-safety planners to remember their major temporary events when leveraging social data in population density, as long as it is an area where the public has used mobile social data at some point during the social data collection period. However, without geotagged social data and only using census grid data, the population density does not show any users and incorrectly shows no need for public-safety broadband coverage.

Geotagged mobile social data should be integral to an overall coverage objective area and may be specified down to a county or tribal area. Data can be determined from a public-safety mission profile template synthesis and include census data, terrain, clutter, location-specific GIS layers, critical infrastructure, required areas of operations, and geocoding of location-specific CAD and 9-1-1 incident data. A baseline coverage objective area could be based on a global composite population density of five or more people per square mile and include both social and census data.

A global composite total population density for areas around Denver and Boulder, Colorado, using geotagged social data shows trails in the mountains west of Denver and Boulder where five or more people clustered at some point during the past five years. These clusters would not show up in the census data alone as a potential coverage area for public safety.

Such a coverage objective target area for the eNodeBs of an LTE public-safety network will encompass most of the coverage target areas of operations including developed or populated areas, roads, critical infrastructure, tribal areas and high-crime incident areas, based on CAD data, where direct loss of life can occur. However, there may be islands of public safety area of operations, such as a stretch of secondary road or a power station, outside of the population density shape file that needs to be part of the overall LTE coverage objective area.

Once coverage objective areas have been established, the number of LTE devices have been determined, and the required minimum cell edge data throughput for uplink and downlink have been established, social data can also be leveraged to prioritize where LTE eNodeB sites should be placed across the given coverage area. Not leveraging geotagged social data and simply placing LTE eNodeB sites uniformly using uniform user distribution can also harm the selling of excess capacity to secondary and tertiary users of the network pending spectrum- and infrastructure-sharing agreements.

First responders require access to a highly available, reliable, uncongested and interoperable wireless broadband network. However, this network must also be aligned with where the general public is most likely to be at any given time in a 24-hour period for both planned and unplanned public-safety events. Leveraging geotagged social data and weighting it into a global composite population density along with other mission-critical data will ensure wireless communications is available when and where it is most likely to be needed.


Robert Dew has more than 20 years of wireless network systems engineering, deployment and operations experience. He has worked or consulted in technical director, project management, and senior systems engineering positions for carriers, vendors and consulting firms in the United States, Europe, Latin America, the Middle East and Asia. Dew received his B.S.in electrical engineering from Virginia Tech and his M.S. in engineering management from The George Washington University.



 
 
Post a comment
Name: *
Email: *
Title: *
Comment: *
 

Comments

No Comments Submitted Yet

Be the first by using the form above to submit a comment!

Site Navigation

Close