Design and Optimisation of an LTE Mobile Phone Antenna
4G/LTE provides a comprehensive and secure all-IP based mobile broadband solution to all kinds of mobile communication devices. WiMAX, HSPA+ and first-release Long Term Evolution (LTE) have been the dominant technologies in this market. These technologies have been designed for channel bandwidths of 5 to 20 MHz (optionally up to 40 MHz) and offer peak data rates of ~ 100 MBit/S for high mobility devices and 1 GBit/s for low mobility devices.
Challenges that typically influence the design of an LTE antenna for a mobile device are:
- Minimal antenna size and tight integration with other device components.
- Mutual coupling between different antennas have to be minimized.
- Compliance with radiation hazard restrictions have to be maintained.
- Phone dimensions:
- Thickness ~ 1cm (slim phones ~ 0.5cm)
- Width ~ 6cm
- Length ~ 12cm
|Novel dual-port antenna typology|
A thorough investigation lead to the decision to use a dual-port antenna configuration for this design. Dual-port antenna designs have the following advantages that are appealing to the current design :
- Two orthogonal radiating elements are used to achieve pattern diversity.
- No additional neutralization stubs or hybrids are used to provide port isolation.
- Zero separation between elements leads to size reduction, which is required for a compact design.
A novel design is proposed in the current work, featuring symmetric radiating elements to keep the radiation characteristics identical for both elements, while enhancing pattern diversity as mentioned above.
The initial design provided good port-port isolation (S21 < -10dB), even thought the ports are physically connected.
This performance is unfortunately not good enough as low correlation is required, while providing good impedance matching at the same time. The current design is considered to be the correct topology, but has to be optimized for a good combination of low port-port isolation and input impedance matching.
Optimized dual-port antenna typology
The initial design is optimized with a combination of the Particle Swarm Optimization (PSO) and Nelder-Mead (Simplex) methods that are available in FEKO. PSO is a good optimization method for large unknown solution spaces, but takes a rather long time to converge. The Simplex method converges much faster, but success of the method depends on a good starting point for the optimization.
The initial design was optimized in a two-step process:
- PSO was run for a few iterations to find a rough global optimum for the antenna geometry.
- The optimum result of the PSO optimization was then used as starting point for Simplex optimization, which converged to the optimum result quite rapidly from this position.
This optimization process resulted in good matching (S11 < -10dB) as well as acceptable isolation (S21 < -10dB) at the desired center frequency.
|(a) Pre-optimization||(b) Optimized at 2.6 GHz|
|Comparison of impedance matching (S11) and port isolation (S21)|
When considering the surface current distribution in the case where both ports are excited, it becomes clear that although the two elements are connected, there is a clear voltage null between the two ports (isolation).
The phase of the two radiating element currents are also in opposite directions, indicating polarization diversity of the antenna.
|Ports isolation and polarization diversity evidenced by POSTFEKO current distribution|
Modern mobile phones tightly integrate many components, all of which influence the performance of the antenna. It is with these effects in mind that the new antenna was integrated in a model of a phone to test the antenna's performance in its operating environment. The following phone structures were included in the model:
- Plastic casing
- LCD display
|CADFEKO Phone model including LTE antenna|
The dual-port antenna in the phone was investigated in three typical modes of operation:
- Both ports excited: The antenna acts as a dual-feed antenna for MIMO applications.
- Port 1 excited, port 2 terminated in a matched load: Port 1 is transmitting while port 2 is receiving.
- Port 1 terminated in a matched load, port 2 open circuited (high impedance): Either antenna can be receiving a signal, giving the device the option of switching between them based on polarity and strength of the incoming signal.
In all 3 cases the radiation characteristics of the antenna mounted on the phone proved satisfactory.
|Port 1 excited, |
Port 2 open
|Port 1 excited and |
Port 2 excited
|Port 1 excited,
Port 2 short
|Radiation patterns for 3 operating states of dual-port antenna mounted on phone|
As a final test, the phone was simulated in operating state 1 (both ports excited) in close proximity to a human head, as the phone would typically be used. The human head acts as a large dielectric load, which can significantly influence the radiation characteristics of the phone. This simulation showed that for far-field radiation patterns both azimuth (horisontal) and elevation (vertical) scan patterns for the phone in proximity to a human head are still acceptable.
|Elevation (Phi = 0º)||Azimuth (Theta = 90º)|
|Phone radiation patterns in proximity of a human head|
SAR Compliance Testing
Does the phone result in acceptable energy absorption by users of the phone?
- Define SAR
- FCC regulations for SAR, Europe (10g cube) vs. US (1g cube)
A common requirement for mobile phones is that energy dissipated in the head of the user has to be below certain limits. These limits are referred to as the Specific Absorption Rate (SAR) of the phone and is measured in Watt per kilogram tissue (W/kg). SAR is measured in three ways:
- Whole body average, which is the total amount of energy dissipated in the human body, expressed as an average of the total mass of the human body.
- 10g cube localized peak, which is the highest level of energy dissipation in any 10g cube of tissue in the human body.
- 1g cube localized peak, which is the highest level of energy dissipation in any 1g cube of tissue in the human body.
Different legislative bodies apply different limits that manufacturers of mobile phones have to adhere to, e.g.:
- European regulators require a 10g cube localized peak SAR < 2 W/kg.
- United States (FCC) regulations require a 1g cube localized peak SAR < 1 W/kg.
Mobile phones typically control the transmitted power of the device dynamically to transmit as little power as necessary for a good communication link. As such, average levels of transmitted power is likely much lower than the maximum possible transmitted power of the device. It then also follows that a SAR investigation at the maximum transmitted power of the device is pessimistic and may serve as a good "worst case" investigation. Assuming a maximum power transmission of 2 W, the novel mobile phone antenna presented here conforms to both European and United States SAR regulations and the phone may therefore be distributed in both regions.
|1g cube localized peak||10g cube localized peak|
|"Worst case" SAR levels for 2 Watt transmitted power|
A novel design for a mobile antenna was presented here, satisfying all the requirements for modern mobile phones. It was also demonstrated how FEKO was applied in the design process, from design prototyping and optimization to integration in a handset and compliance testing of the handset.
In summary, FEKO proved to be an invaluable tool through the entire design cycle of mobile phone antennas.
Original publication of this work in Microwave Journal, March 2012: "Compact Antenna for MIMO LTE Mobile Phone Applications."
|||Rao, Q., and Wang, D., "A Compact Dual-Port Antenna for Long-Term Evolution Handheld Devices", IEEE Transactions on Vehicular Technology, Vol. 59, No. 3, March 2010|