Design and Optimization of an LTE Mobile Phone Antenna

In this white paper an LTE antenna is designed for use in a mobile phone, using a dual-port configuration. Using FEKO, the antenna is also optimized for minimal reflection co-efficient at both ports and cross-coupling between the ports as low as possible.


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

Dual-Port Antenna

Initial design

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 [1]:

  • 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.
Novel dual-port antenna typology

Novel dual-port antenna typology

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. Substrate: FR4, Thickness: 5mm, Dielectric constant: 4.8, Loss tangent: 0.017

The initial design was optimized in a two-step process: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.

  1. PSO was run for a few iterations to find a rough global optimum for the antenna geometry.
  2. 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.

Comparison of impedance matching (S11) and port isolation (S21)

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.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).

Ports isolation and polarization diversity evidenced by POSTFEKO current distribution

Ports isolation and polarization diversity evidenced by POSTFEKO current distribution

Phone Integration

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
  • Battery
  • LCD display

CADFEKO Phone model including LTE antenna

Both ports excited: The antenna acts as a dual-feed antenna for MIMO applications.The dual-port antenna in the phone was investigated in three typical modes of operation:

  1. Both ports excited: The antenna acts as a dual-feed antenna for MIMO applications.
  2. Port 1 excited, port 2 terminated in a matched load: Port 1 is transmitting while port 2 is receiving.
  3. 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.

Radiation patterns for 3 operating states of dual-port antenna mounted on phone

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 (horizontal) and elevation (vertical) scan patterns for the phone in proximity to a human head are still acceptable.

Phone radiation patterns in proximity of a human head

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.

"Worst case" SAR levels for 2 Watt transmitted power

“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.


[1] 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

Antenna Design for 5G Communications

With the rollout of the 5th generation mobile network around the corner (scheduled for 2020 [wiki/5G]), technology exploration is in full swing. The new 5G requirements (e.g. 1000x increase in capacity, 10x higher data rates, etc.) will create opportunities for diverse new applications, including automotive, healthcare, industrial and gaming. But to make these requirements technically feasible, higher communication frequencies are needed. For example, the 26 and 28 GHz frequency bands have been allocated for Europe and the USA respectively – more than 10x higher than typical 4G frequencies. Other advancement will include carrier aggregation to increase bandwidth and the use of massive MIMO antenna arrays to separate users through beamforming and spatial multiplexing.

Driving Innovation Through Simulation

The combination of these technology developments will create new challenges that impact design methodologies applied to mobile and base station antennas currently. Higher gain antennas will be needed to sustain communications in the millimeter wavelength band due to the increase in propagation losses. While this can be achieved by using multi-element antenna arrays, it comes at the cost of increased design complexity, reduced beamwidth and sophisticated feed circuits.

Simulation will pave the way to innovate these new antenna designs through rigorous optimization and tradeoff analysis. Altair’s FEKO™ is a comprehensive electromagnetic simulation suite ideal for these type of designs: offering MoM, FEM and FDTD solvers for preliminary antenna simulations, and specialized tools for efficient simulation of large array antennas.

Beam antenna pattern

Mobile Devices

In a mobile phone, antenna real estate is typically a very limited commodity, and in most cases, a trade-off between antenna size and performance is made. In the millimeter band the antenna footprint will be much smaller, and optimization of the antenna geometry will ensure the best antenna performance is achieved for the space that is allocated, also for higher order MIMO configurations.

At these frequencies, the mobile device is also tens of wavelengths in size and the antenna integration process now becomes more like an antenna placement problem – an area where FEKO is well known to excel. When considering MIMO strategies, it is also easier to achieve good isolation between the MIMO elements, due to larger spatial separation that can be achieved at higher frequencies. Similarly, it is more straightforward to achieve good pattern diversity strategies.

Base Station

FEKO’s high performance solvers and specialized toolsets are well suited for the simulation massive MIMO antenna arrays for 5G base stations. During the design of these arrays, a 2×2 subsection can be optimized to achieve good matching, maximize gain and minimize isolation with neighboring elements –a very efficient approach to minimize nearest neighbor coupling. The design can then be extrapolated up to the large array configurations for final analysis. Farming of the optimization tasks enables these multi-variable and multi-goal to be solved in only a few hours. Analysis of the full array geometry can be efficiently solved with FEKO’s FDTD or MLFMM method: while FDTD is extremely efficient (1.5 hrs for 16×16 planar array), MLFMM might also be a good choice depending on the specific antenna geometry.

The 5G Channel and Network Deployment

The mobile and base station antenna patterns that are simulated in FEKO, can used in WinProp™ for high-level system analysis of the 5G radio network coverage and to determine channel statistics for urban, rural and indoor scenarios.

WinProp is already extensively used for 4G/LTE network planning. However, the use cases for 5G networks will be even more relevant largely due to the different factors that occur in the millimeter band. These include higher path loss from atmospheric absorption and rainfall, minimal penetration into walls and stronger effects due to surface roughness.

In addition to being able to calculate the angular and delay spread, WinProp also provides a platform to analyze and compare the performance of different MIMO configurations while taking beamforming into account.

The Road to 5G

While some of the challenges that lie ahead to meet the 5G requirements may still seem daunting, simulation can already be used today to develop understanding and explore innovative solutions. FEKO offers comprehensive solutions for device and base station antenna design, while WinProp will determine the requirements for successful network deployment.

Request more information on FEKO

Tips & Tricks: June 2017

T & T 1236 – solidThinking Inspire – Keyboard Shortcuts T & T 1237 – OptiStruct – PARAM,MAXRATIO to detect the singularity T & T 1238 – OptiStruct – Normalized Constraint Violation Plot by Optimization run T & T 1239 … Read More