Testbed for Prototype C-Band Airport Surface Wireless Network

/
International Civil Aviation Organization
WORKING PAPER / ACP-WGF15/WP25
26/05/06

AERONAUTICAL COMMUNICATIONS PANEL (ACP)

FOURTEENTH MEETING OF WORKING GROUP F

Cairo, Egypt 7 – 13 June 2006

Agenda Item 5: / AM(R)S spectrum requirements (Future Communications System (FCS))

Testbed for Prototype C-band Airport Surface Wireless Network

(Presented by Robert Kerczewski, NASAGlennResearchCenter)

Prepared by Phil Neumiller, Dana Hall, and Steve DeHart of Sensis Corporation

and Robert Kerczewski of the NASAGlennResearchCenter

SUMMARY
As part of a Next Generation Air Transportation System testbed, NASA is working with the Sensis Corporation and ClevelandHopkinsInternationalAirport to deploy and test a prototype wireless airport surface network, intended for operation in the 5091-5150 MHz band, beginning in June 2006.
ACTION
The meeting is invited to note these activities and provide comments.

Abstract

A collaboration between NASAGlennResearchCenter, the Sensis Corporation, and the ClevelandHopkinsInternationalAirport is resulting in the development of a multi-site testbed for advanced communications/navigation/surveillance concepts for the Next Generation Air Transportation System (NGATS). A key feature of this testbed will be the prototyping of an airport surface wireless network for ATC/AOC communications needed to meet future NGATS requirements for more efficient and safe airport operations. This wireless network is intended to operate in the 5091-5150 MHz band, which provides the necessary bandwidth while meeting the short transmission distance restrictions required to meet interference limits. The paper describes the testbed concept, design aspects of the airport surface wireless network, and deployment and test plans.

1 - Introduction

Increasing the safety and efficiency of airport operations is a critical aspect of enabling the Next Generation air Traffic System (NGATS) envisioned by the US Joint Planning and Development Office (JPDO). Increased requirements for airport weather and surveillance sensors, navigation equipment, and communications for traffic management activities make a modern wireless communications network is necessary.

Other WGF-15 working papers discus the communications and bandwidth requirements for an airport surface network. The 5091-5150 MHz (i.e. microwave landing system (MLS) extension band) provides the required bandwidth for an application which is ideally suited for a band in which interference requirements dictate that only short distance communications links can be implemented. In addition, such short distance links ensure that the frequency band can be reused multiple times at different airport locations.

The Eurocontrol/FAA Future Communications Study Phase I technology assessment concluded that the IEEE 802.16 standard is a good choice for an airport surface wireless network based on commercial standards. 5091-5150 MHz channel sounding measurement campaigns conducted at several airports, described in another working paper, indicate acceptable propagation performance in a dynamic airport environment. What remains next is to develop and evaluate a prototype system to identify optimum system design parameters.

The NGATS CNS National Testbed will enable the development, testing and evaluation of many concepts and technologies necessary to enable the JPDO NGATS vision in the airport and terminal area environment. A key aspect will be an airport surface wireless network testbed enabling the development and evaluation of a prototype 5091-5150 MHz wireless network based on IEEE standards in an active airport environment. Initial laboratory and field testing has already taken place, and initial deployment of the testbed begins in the summer of 2006.

This working paper will present an overview of the NGATS CNS testbed concept, and describe the 5091-5150 MHz airport surface wireless network prototype approach, deployment schedule and test plans.

2 - NGATS CNS National Testbed Concept

NASA’s GlennResearchCenter and the Sensis Corporation have defined and are implementing a Test Bed to explore technologies and systems appropriate to the transformation of the Nation’s air transportation system. Responding to and shaped by the concepts formulated by the federal and industry Joint Planning and Development Organization, the NGATS National Test Bed will be instantiated at three ClevelandOhio area airports, at the nearby NASA Glenn campus, and at Sensis near SyracuseNew York. The three selected airports, Hopkins International, Burke Lakefront, and Lorain County Regional (Figure 1), are representative of the classes of airports across the United States today. An advanced cooperative surveillance system integrated by wireless communications will be installed at these fields during the summer of 2006. This “infrastructure” will then be used to quantitatively evaluate leading JPDO NGATS concepts and alternative solutions. Examples of such applications are approaches for remote control of airports via staffed facilities perhaps complemented by automated sequencing and other advisory tools, information access and sharing through a common NAS service oriented architecture, and the procedures and automation by which aircraft in flight and on the airport surface will be able to negotiate travel trajectories. Applications such as these require a robust, highly reliable, and operationally cost-effective surveillance and communications infrastructure. The remainder of this paper focuses on the characteristics of the portion of the surface communications infrastructure central to the control and monitoring of aircraft.

Figure 1 – NGATS-CNS Testbed Sites

3 - Communications Infrastructure

The communications infrastructure consists of four elements (see Figure 2). Each airport will have a surface wireless network. This network is broken down into three functional entities. The “Blue” network is a wireless mesh backbone consisting of wireless nodes distributed across the airport surface at strategic locations. The “Red” network is a collection of wireless, local access nodes, collocated with Blue network nodes, analogous to a cellular network. The Red network provides network access to both fixed and mobile users that cannot easily be physically connected to Blue nodes. The “Black” network is an administrative network that provides remote, council port access to both Blue and Red network nodes for the purpose of remote configuration, reset, firmware upgrade, experiment administration, etc. This can be done from a single central location within the testbed. The Black network is primarily to facilitate testbed activities including experiments, test administration and reconfiguration without having to physically access Blue and Red node hardware.

The wireless networks at all three airports are administered from a Network Operations Center (NOC) located at NASA Glenn. A router at the NOC serves as a common traffic termination point for all traffic from the airport networks. The Burke Lakefront and Lorain County locations are linked to the NOC via a wired Wide Area Network (WAN) supplied by commercial network provider(e.g. Sprint, AT&T, or Verizon, etc.). The WAN also provides a link to Sensis Corporation in Syracuse, NY. Due to close proximity, Cleveland Hopkins airport is linked to the NOC via two wireless gateways which are part of the Hopkins Blue network.

Figure 2 – Testbed Communications Infrastructure

4 - C-BandAirportSurface Wireless Network Prototype Testbed

The wireless network testbed will evaluate thecapabilities, performance and limitations of an airport surface wireless network infrastructure, operating in the 5091-5150 MHz band, in a real operational airport environment. The network will also serve as an infrastructure for the development, testing and evaluation of many NGATS concepts and technologies.

The wireless infrastructure utilizes commercial off-the-shelf (COTS) 802.11a wireless technologies, modified to operate in the desired 5091-5150 MHz band, during the initial prototype phase of the testbed. This phaseof the testbed deployment and evaluation focuses on the performance of the Blue backbone network.

The Blue network is a wireless mesh network. The mesh architecture provides a resilient network that adapts to different link performance conditions and node possible hardware failures. When the network is first activated, each node establishes a link with all nodes that are detectable from that site. Each of these links is evaluated and ranked base on performance parameters and preference criteria set by the network administrator. From this link and rank information, each node determines primary and secondary (backup) network paths for all possible traffic destinations. In the event of a hardware failure or link degradation, the network routes traffic via the appropriate secondary network paths.

The physical wireless network architecture is based on network node sites strategically placedto provide both coverage of the airport surface and reliable connectivity between Blue network nodes. Each node site consists of a Blue network node radio, a Black network radio, an Ethernet switch, a single board computer (SBC) and associated power supplies, surge protection and enclosure hardware. The SBC supports site monitoring, data collection, test data generation and other various functions required for network experiments and detailed testing.The Ethernet switch provides connectivity between the Blue radio different data sources located at the site. These include Red network radio(s), local sensors and one of the SBCdata ports for test traffic. The Black radio interfaces with the Blue radio console port and the SBC remote site control and monitoring from the NOC at NASA Glenn.

5 - Laboratory and Syracuse Hancock Airport Testing

Network requirements were developed by evaluating the traffic requirements of network users, sensors and applications. COTS hardware products were selected based on these criteria. The next step was to evaluate the network architecture and components under controlled conditions. Two major performance categories were determined for testing. These were network behavior & performance, and RF propagation & link performance. Testing performed at NASA Glenn tested an initial three-node configuration. Testing at Sensis Corporation laboratories exercised the planned Phase I hardware in a six-node version in a laboratory environment, and tested link performance sensitivity at the SyracuseHancockAirport.

5.1 - NASA Glenn Network Tests

The NASA Glenn Research Center has implemented an experimental three-access point prototype surface wireless network interconnecting Local Area Networks (LANs) that reside in three buildings at the Glenn campus, known as buildings 7, 55, and 333. This network is currently configured to provide optimum data rates between buildings 7 and 55 by utilizing their wireless link to building 333, which bridges the aforementioned buildings as it provides wireless service for its own LAN. The experimental network is being used to investigate network for later testing in the NGATS-CNS testbed.

Glenn’s experimental network utilizes the IEEE 802.11g protocol specification, operating at 2.4 GHz. Because 802.11g is a half duplex protocol, throughput is reduced by 50% (and latencies doubled) with each single access point wireless hop. This throughput reduction can be mitigated by the placement of a second wireless access point (operating on a different channel) within building 333.

The Glenn experimental network has tested voice-over-IP (VoIP) applications to exercise the network and gather performance data. Preliminary testing of a synchronous circuit over VoIP connected via a wireless link demonstrated maintenance of synchronization of a circuit at a 14.4Kbps data rate while sustained data rates of 1Mbps were being utilized for other traffic.

5.2–Sensis Corporation Labopratory and field testing

Network performance is being tested in the laboratory with a scaled down, six node version of the network. The electronics from six node products were enclosed in RF isolation enclosures allowing the network to be implemented in a small area of the laboratory as shown in Figure 3. The node product selected utilizes a 360° array of eight 45° sector antennas connected to an eight way RF switch. During normal network operation, the RF switch selects the optimum antenna for the appropriate link path for a given transmission interval. In the lab setup, three of the switch antenna ports were connected to RF ports on the wall of the isolation enclosure. Mesh network “links” were created by connecting different ports of different nodes with coax cables and attenuators. The attenuators are computer controlled, allowing for the simulation of link fading and outages. Other computers connect to the traffic ports on each node to simulate various traffic load profiles. Network testing includes mesh self forming & healing, latency, capacity, Quality of Service functions, network monitoring capabilities, etc.

The 5091-5150 MHz propagation environment on the airport surface provides significant challenges. Large metal structures can create significant multipath reflectors. Moving aircraft and large vehicles can create time varying multipath and/or cause link obstruction. These effects have been studied and characterized by OhioUniversity and are the subject of another WGF-15 working paper.

Avoiding these propagation “hazards” will have a significant affect on the overall performance of the network. Node placement, both position and height above ground level AGL play a significant role in link susceptibility and performance. To get a preliminary understanding of the sensitivity of the link performance to aircraft link obstruction, a test link was set up at SyracuseHancockAirport. Figure 4 shows the link path was chosen. The path crossed the most frequently used runway and taxiway.

The link hardware configuration is shown in Figure 5. A pair of 5 GHz wireless nodes were mounted at the two link termination sites. Each site had a computer to generate and evaluate test traffic carried over the link. A 900 MHz link was also set up to control the remote site. Measurements in this test were made at 5.3 GHz (unlicensed, middle UNII band) as a 5091-5150 MHz experimental license was not available at the time of the test. This is a 4% offset in frequency and has minimal effect on link behavior.

Link parameters and performance were measured as various sized aircraft and vehicles passed through the link path. Variations in received signal strength, packet loss rate were recorded vs. aircraft/vehicle position. These measurements were repeated for different antenna heights. As expected, optimum node locations are those that offer good antenna height (within airport restrictions) and locations that support primary link paths that avoid aircraft movement areas. This is the strategy used in selecting the node site locations for the testbed. The testbed will offer the opportunity to measure link behavior under a wider range of conditions and aircraft types.

Figure 3 – Sensis Lab Test Configuration

Figure 4 – SyracuseHancockAirport Test Links and Locations

6 - Next Steps in Testbed Development

There several ongoing tasks to be completed before deploying the wireless testbed at the three Clevelandarea airports. In the laboratory, test automation scripts and performance tools are being completed. Test and configuration automation will facilitate consistent and efficient system testing. Applications for siting permits and experimental licenses are being prepared for submission. Site components and hardware will be ordered and individual site “kits” will be assembled at Sensis before shipment to Cleveland for installation.

Network deployment and integration will be done in stages. First the Network Operations Center (NOC) will be assembled and integrated at the NASA Glenn site. Next, the commercial WAN will be activated and integrated with the router and network severs at the NOC. With this base infrastructure in place, the wireless networks at each airport can be activated and integrated into the total communications infrastructure. From this point, the full spectrum of network testing can be performed ranging from link performance in an active runway environment to network testing including basic network performance (capacity, latency, packet loss), mesh performance (mesh forming, stability, failover) and QOS (rate control, packet prioritization, granularity).

Figure 5 – Outdoor Test Configuration

When the full spectrum of network testing and experiments is finished, the network can then be made available to support the broader range of NGATS experiments planned for the testbed. 802.16wireless network technologies will be procured, tested and phased into the airport surface network infrastructure and improvements gauged against current performance.

7 - Testbed Deployment and Testing Schedule

The deployment, integration and field testing will role out as follows. Laboratory testing and tool development will continue through June 2006. Site kitting and major subassemblies will be built during July 2006. NOC and WAN network integration will occur late July and extend into August 2006. The airport wireless networks will be deployed in August with network testing extending into September 2006.

8 - Summary/Conclusions

Increased requirements for airport sensors, communications, surveillance and traffic management demand a flexible, capable and reliable communications infrastructure. An airport surface wireless network using advanced wireless network technology is a strategic part of meeting this demand. The 5091-5150 MHz extended MLS band, when allocated for ATC/AOC safety critical applications will provide the spectrum bandwidth required for this type of network. As a part of the NGATS CNS National Testbed, the experimental airport wireless network will provide the environment necessary to develop this communications infrastructure.

Page 1 of 10