Skip to Main Content
In a world where mobile connectivity has become a need, aircraft passengers still have limited access to communication services during the flight. Both industries and research communities have tried to solve this problem by applying a number of wireless access technologies within the cabin. However, these services are not currently available in commercial aircrafts, where the users have to spend hours isolated from the rest of the world. The universal mobile terrestrial system (UMTS) has been considered as a candidate wireless access network for providing mobile connectivity. Since UMTS is not a near range technology, it was always assumed that one Node B is enough to cover the whole cabin. However, if only one UMTS Node B is adopted a relatively large transmission power would be required to guarantee an acceptable quality of service. As a consequence, this will interfere with on-board instrumentation and control transmission mechanisms. Moreover some of the signal may radiate outside the cabin possibly interfering with the terrestrial UMTS networks. This type of network topology has the disadvantage of having only one single point of failure and therefore the connection will not be available if the propagation path between the UMTS Node B and the receiving point are obstructed. This paper tries to solve these problems by applying a number of UMTS Node Bs strategically located around the aircraft, taking a typical configuration of the Airbus A340-600. The optimized network can be derived through simulation where the cabin is modeled as a set of surfaces that form the aircraft fuselage, seats and stowage bins. The dielectric properties of the media, the allowed antenna locations and the antenna patterns were stored in text files and imported in the simulation environment for processing. A number of transmitting antennas are then placed at the available antenna locations and assigned the transmitting power of the UMTS Node B. A three-dimensional ray launching algorith- m based on geometric optics (GO) was adopted to derive the propagation paths of the rays launched by the transmitting antenna. A ray sent out by the transmitter travels in free space until it impinges on a surface. The impinged point is then used to transmit one or two rays, whose directions are derived according to Fermat' s principle while the transmitting power of each ray are derived by multiplying the power of the impinged ray to Fresnel's reflection and transmission coefficients. This process goes on for each ray until the propagation power of all the launched, reflected and refracted rays are below a certain threshold. This process is repeated several times, each time deriving the coverage obtained by the considered network topology. Once an acceptable number of iterations are computed, the topology which provides the best compromise between the number of Node Bs and transmission power levels is selected. The low transmission power required by the proposed solution ensures that the interference on the aircraft signaling infrastructure is minimal. Moreover, the Node Bs can be connected via point-to-point wireless links and therefore the extra weight incurred is minimal.