EP3381132A1

EP3381132A1 - Traces between phase array antenna and radio frequency integrated circuit in mm wave systems

Info

Publication number: EP3381132A1
Application number: EP16921313.9A
Authority: EP
Inventors: Krishna S. Gomadam; Djordje Tujkovic; Sanjai Kohli
Original assignee: Facebook Inc
Current assignee: Meta Platforms Inc
Priority date: 2016-11-10
Filing date: 2016-11-10
Publication date: 2018-10-03
Also published as: WO2018089008A1; EP3381132A4; CN109644031B; CN109644031A

Abstract

Systems and associated methods for improved beamforming of the phase array antenna are disclosed herein. In one embodiment, a communication system for wireless signals has a phase array antenna having a plurality of individual antennas and a plurality of electrically conductive traces. The individual traces electrically connect corresponding individual antennas with a transmitter. The lengths of individual traces Ti, Tk satisfy equation Abs ((Ti-Tk) mod (λ)) < λ/B, where λ is a wavelength of the wireless signal and λ/B is a fraction of λ.

Description

TRACES BETWEEN PHASE ARRAY ANTENNA AND RADIO FREQUENCY INTEGRATED CIRCUIT IN MM WAVE SYSTEMS

TECHNICAL FIELD

[0001] The present technology is generally related to the field of phase array antennas, more specifically to the field of reducing power losses and improving signal to noise ratio (SNR) for phase array antennas.

BACKGROUND

[0002] An antenna (e.g., a dipole antenna) typically generates radiation in a pattern that has a preferred direction, i.e., the generated pattern is stronger in some directions and weaker in other directions. When receiving electromagnetic signals, the antenna has the same preferred direction, i.e., received signal is strongest in the preferred direction. Therefore, signal quality (e.g., signal to noise ratio) can be improved by aligning the preferred direction of the antenna with a target receiver or a source of signal. However, it is often impractical to physically reorient the antenna and/or the exact location of the target may not be known. To overcome some of the above shortcomings of the antenna, a phase array antenna can be formed from a set of antennas to simulate a large directional antenna. An advantage of the phase array antenna is its ability to transmit/receive signals in a preferred direction (i.e., its beamforming ability) without physically repositioning or reorienting the antenna.

[0003] Figure 1 is a schematic illustration of a phase array antenna system 10 according to prior art. The illustrated system has a phase array antenna 14 that includes four individual antennas 14i that are set apart by a half wavelength (λ/2) of the transmitted signal. A transmitter 12 generates signals for the phase array antenna 1 0. The transmitter 12 includes a modulator that receives two inputs (a baseband signal and a carrier oscillator) and outputs a modulated radio frequency (RF) signal. For example, a baseband signal having a relatively low frequency can be modulated by a carrier oscillator having a relatively high frequency to produce the modulated RF signal. The resulting modulated RF signal is transmitted through a beamformer 20 that adjusts amplitude and phase of the RF signal by applying an amplitude adjustment (ai) and a phase shift (θί) to the RF signal. A combination of the amplitude and phase adjustment for each individual antenna 14i is called a complex weight (wi) for that antenna. Since the resulting adjusted RF signal (i.e., adjusted by applying the complex weight) is generally a low power signal, power amplifiers 38i amplify RF signal that leaves the beamformer 34. Amplified RF signals arrive at the individual antennas 14i, and are transmitted to surrounding space as a wireless signal. In the illustrated example in Figure 1 , the wireless signal is transmitted in a direction D, i.e., a front F of the wireless signal makes an angle a (angle of antenna or AoA) with respect to the plane of the antennas 14i. The desired direction D of the wireless signal can be achieved by, for example, programming the beamformer 20 such that Θ1 > Θ2 > Θ3 > Θ4 by appropriate ΔΘ, such that the front F makes angle a (AoA) with the plane of the antennas 14i. Analogously, if the phase array antenna 14 is a receiver and a source of RF signal (i.e., a transmitter) is located in at AoA D', the distribution of 9i's can be adjusted such that the phase array antenna has a maximum sensitivity in the direction D'.

[0004] However, in addition to the desired, controlled changes to phase θί of the individual antennas 14i, the system may introduce undesired changes to phase θί. For example, high frequency RF signals (e.g., GHz range) are characterized by short wavelengths (e.g., mm range wavelegths). Consequently, even a relatively small difference between the length Ti of transmission lines 19i (e.g., mm or sub-mm differences) may result in appreciable deviation from the desired distribution of 9i's, which, in turn, causes errors in the AoA, loss of sensitivity, spurious/undesired lobes around the AoA, and other issues. One approach to minimize these issues is described in relation to Figure 2 below.

[0005] Figure 2 is a schematic view of a phase array antenna system 200 in accordance with prior art. The illustrated phase array antenna 14 includes seven individual antennas 14i spaced apart by half wavelength (λ/2) of the transmitted signal. The transmitter 12 generates RF signals that are routed to the antennas 14i through traces 29i. The signals in the individual traces 29i can be phase shifted to produce desired AoA of the outgoing wireless signal. As explained above, differences in the length Ti among individual traces 29i can introduce undesirable variations in the phase shifts (9i's) of the individual antennas 14i, especially for the signals in the GHz frequency range. Therefore, with the illustrated conventional system, all traces 29i have uniform length to eliminate/minimize the unwanted variability in the phase of signal transmitted through the traces. However, such uniform length must necessarily correspond to at least the distance from the transmitter 12 to the peripheral antennas 14i. As a result, the traces 29i that connect the transmitter 12 with the centrally located antennas 14i are longer than necessary, which causes additional signal power loss in these traces. Accordingly, it would be advantageous to provide traces that connect the transmitter/receiver with the antennas such that the changes of phase θί caused by the traces are minimized, while not introducing an undesirable power loss.

BRIEF DESCRIPTION OF THE DRAWINGS

[0006] Figure 1 illustrates a phase array antenna 10 according to prior art.

[0007] Figure 2 is a schematic view of a phase array antenna 200 in accordance with prior art.

[0008] Figure 3A is a schematic view of a phase array antenna 300 in accordance with an embodiment of the present technology.

[0009] Figure 3B is a simplified diagram of the trace lengths in accordance with an embodiment of the present technology.

[0010] Figure 4 is a schematic illustration of a tiled phase array antenna 400 in accordance with an embodiment of the present technology.

[0011] Figures 5A and 5B are schematic illustrations of a transceiver in accordance with an embodiment of the present technology.

[0012] Figure 6A is a graph of signal gain in dB for traces of equal and unequal length when the trace loss is 0.25 dB/cm.

[0013] Figure 6B is a graph of signal gain in dB for traces of equal and unequal length when the trace loss is 0.5 dB/cm.

DETAILED DESCRIPTION

[0014] The following disclosure describes various embodiments of systems and associated methods for minimizing unwanted variations in signal phase (Θ) of the individual antennas in a phase array antenna. Most phase array antennas require precisely controlled differences between the signal phase of the individual antennas to produce desired directivity (i.e., angle of antenna or AoA) of the wireless signal of the phase array antenna. The unwanted variations in Θ cause errors and/or increase noise in the directivity of the phase array antennas. To achieve this coherence, conventionally the traces to the elements of the phase array were made to have equal length, dictated by the length of a trace to the most distant element of the array, i.e., all the traces had to be as long as the longest trace. As a result, the conventional approach causes undesired power losses for the centrally located traces that could have used shorter traces.

[0015] In some embodiments of the present technology, output pins of a Radio Frequency Integrated Circuit (RFIC) chip are connected by traces to individual antennas of the phase array antenna. The length of the traces is set such that a difference in the trace lengths for any two traces corresponds to a multiple of signal wavelength. Thus the traces leading to the elements (i.e., individual antennas) located centrally within the phase array antenna are shorter, and the traces to the peripheral and more distant elements of the phase array antenna are longer. However, sometimes the phase array antenna operates at variable signal wavelength, and the difference in trace length cannot be set at a precise multiple of wavelengths. Instead, a variation in the trace length can be permitted up to a certain value, e.g., up to a quarter wavelength or up to an eighth wavelength (i.e., λ/4 or λ/8). The resulting error can be corrected by utilizing a beamforming that takes into account the trace length mismatch, for example, an optimal beamformer that uses the best DFT vector.

[0016] Figure 3A is a schematic view of a phase array antenna system 300 in accordance with an embodiment of the present technology. The illustrated embodiment of a phase array antenna 14 includes nine antennas 14i that are spaced apart by λ/2, but other number of antennas 14i and other distances between the individual antennas are also possible. In general, when the distance between the individual antennas is about λ/2, the undesirable side lobes are eliminated for the directivity of the phase antenna array. Traces 39i that connect a transmitter 32 (e.g., an RFIC chip) to the individual antennas 14i can be selected such that a difference in lengths Ti between any two traces 39i is a whole multiple of the signal wavelength λ. Such a condition can be expressed as: (Ti-Tk) mod (λ) = 0 (Equation 1 )

where "mod" is a modulo operand that calculates the remainder (R) after dividing one number by another. When modulo operand results in zero, the argument to the left of the operator is a whole multiple of the argument to the left of the operand. In practical applications, the remainder that is sufficiently close to zero may be considered zero. For example, in some embodiments of the present technology the remainder that is smaller than λ/10 or λ/100 may be considered close to zero or being substantially zero. In some embodiments, the transmitter 32 (e.g., the RFIC chip) can include a beamformer (not shown for simplicity).

[0017] However, if λ is not constant (e.g., λ = 5.1 , 5.0, 4.8, 4.6 mm for signal with frequency 58.32, 60.48, 62.64, 64.80 GHz, respectively) then (Ti-Tk) mod (λ) will be nonzero for at least some operational A's. Therefore, in some embodiments of the present technology, trace lengths Ti can be selected such that the value of remainder R does not exceed defined limits. For example, the value of remainder R can be defined as a fraction of wavelength λ that is not to be exceeded for any pair of traces Ti and Tk, as in Equation 2 below.

Abs ((Ti-Tk) mod (λ)) < λ/Β (Equation 2) where Abs is an absolute value operator and B is a number that defines what fraction of λ is used as the upper limit of the expression on the left hand of Equation 2.

[0018] Figure 3B is a simplified diagram of the trace lengths in accordance with an embodiment of the present technology. In the illustrated embodiment, a trace 39-1 has length T1 and a trace 39-2 has a length T2. The difference between T1 and T2 is λ/Β, where B can be, for example, 2, 4 or 8. Therefore, traces 39-1 and 39-2 satisfy Equation 2 in limit (i.e., if the difference in trace length is up to λ/Β). Furthermore, a trace 39-3 has a length N^* λ+ λ/Β, where N is a whole number, i.e., N^* λ is a whole multiple of λ. Therefore, a pair of traces 39-1 and 39-3 will also satisfy Equation 2 in limit. A person having ordinary skill would recognize that other pairs of illustrated traces 39-1 to 39-4 will also satisfy Equation 2. Other combinations of trace length are also possible, for example, T3 being up to λ/Β longer than T1 ; T3 being up to Ν^*λ- λ/Β shorter than T1 ; and T4 being up to Ν^*λ+ λ/Β longer than T1 . Four traces are shown in Figure 3B for illustration, but other number of traces and the corresponding antennas is also possible. [0019] When the pairs of traces 39i having lengths Ti, Tk satisfy Equation 2, a degradation of a beamformer gain for different wavelengths λ is limited. An example comparison of the degradation of the beamformer gain versus an ideal beamformer gain of the phase array antenna is shown in Table 1 below.

Table 1

[0020] As shown in Table 1 , the degradation of the gain of the phase array antenna is limited by the restrictions placed on the differences in in trace lengths defined by Equation 2. For example, for a 2 bits phase (i.e., the phase of the signal changes by 2TT/4 increments), the ideal gain (i.e., modulus of zero) of the phase array antenna is 14.7 dB, whereas for the modulus of λ/2 the gain is 10.5 dB. Therefore, a degradation of the gain is 4.2 dB. However, if the modulus is limited to, e.g., λ/8 then the gain of the phase array antenna is 14.5, i.e., the degradation of the gain is only 0.2 dB when compared to the ideal gain of that particular phase array antenna. For a 5 bits phase (i.e., the phase of the signal changes in 2ττ/32 increments), the ideal gain of the phase array antenna is 15.55 dB, whereas for the modulus of λ/2 the gain is 1 1 .2 dB, and for the modulus of λ/8 the gain is 15.35 dB. Therefore, by reducing the limits of the modulus in Equation 2, the degradation of the gain of the phase array antenna is reduced from 4.35 dB (i.e., 15.55 - 1 1 .2 dB) to 0.2 dB (i.e., 15.55 - 15.35 dB). For many practical applications, a degradation of the gain of less than 1 dB can be smaller than a degradation caused by a power loss associated with having identical length for all traces that connect the RFIC with the elements of the phase array antenna. [0021] The above discussion of Equation 2 and Table 1 relates to changes in phase Θ of the wireless signal. However, in general, in beamforming of the phase array antenna 14 both the phase and amplitude of each antenna element 14i are controlled. In some embodiments, controlling both phase and amplitude can adjust side lobes and can improve the AoA of the phase array antenna 14 more precisely than the control of phase alone. The combined relative amplitude ai and phase shift θί for each element of the antenna is a complex weight wi (for the ith individual antenna 14i). A beamformer 20 for the phase array antenna applies this complex weight wi to the signal (i.e., the beamformer shifts the phase and adjusts the amplitude) for each element of the phase array antenna 14. Therefore, an optimal beamformer applies a set of complex weights wi that result in a maximum gain for the wireless signal.

[0022] In general, optimality of the beamforming is preserved as long as the phase offsets generated by the unequal trace lengths can be cancelled. In some embodiments of the present technology, the phase rotation caused by the ith trace having length Ti is -2πΤί/λ for the trace lengths Ti that satisfy Equation 2. Therefore, if a set of the complex weights w'i is the optimal beamformer 20 for the phase array antenna 14 having the traces 39i of equal length, then the optimal beamformer for the phase array antenna 14 having traces of unequal length applies the complex weights:

wi = exp 0^*2πΤί/λ) ^* w'i (Equation 3) where j is a unit imaginary number. The trace length can be calibrated by measuring the relative phases at the antennas when receiving a tone transmitted from broadside. Alternatively, the relative phases can be adjusted and the set of weights that maximize the received power can be determined.

[0023] The complex weights wi of the optimal beamformer can be quantized depending on how many bits of resolution are used for the phase of the signal. For example, for a 2 bits phase discretization, the signal phase changes in 2ττ/4 increments, and the complex weights wi can be quantized using values from the set (0, ττ/2, ττ, 3ττ/2). Similarly, for a 4 bits phase discretization, the signal phase changes in 2ττ/16 increments, and the complex weights wi can be quantized using values from the set (0, ττ/8, ττ/4, 3ττ/8, .... 7ττ/8, π). Other sets for the complex weights wi can also be created depending on how many bits are used for the phase quantization. [0024] Figure 4 is a schematic illustration of a tiled phase array antenna 400 in accordance with an embodiment of the present technology. In general, the systems and methods described with reference to Figure 4 are applicable to high frequency multiple antenna systems. These systems and methods are described below in the context of the variation in the trace length. In the illustrated embodiment, signals from a baseband integrated circuit (BBIC) can be routed through a two-stage arrangement of RFIC's to a set of phase array antennas 440i. Signals from the RFIC3 in the first stage (e.g., a master RFIC) can be routed through sets of traces TRFIC to RFIC's 1 -3 (e.g., a slave RFIC's) which then drive three phase array antennas 440i, but other numbers of RFIC's and other number of stages are also possible. In general, trace lengths are subject to manufacturing variations and/or tolerances that can reduce gain of the tiled phase array antenna 400. As explained above, in some embodiments of the present invention, the trace length from the RFIC to individual elements of the phase array antenna may be different by design, while the additional, manufacturing-induced variability of the trace length is a random variation. Taking into account the manufacturing-induced variability, relationship among the lengths of the traces can be expressed as:

Ti - Tj = Dij + Nij (Equation 4)

where Dij is a known difference (e.g., set by the design) between the length of two traces i and j, and Nij is a random variation introduced by manufacturing errors/tolerances. For example, if the manufacturing errors/tolerances in trace length are about 1 mm and the signal frequency is 58.4 GHz , the resultant variability in the phase of the signal is about +/- 0.39TT. This variability is unaccounted for by the design and corresponds to about 70^Q phase shift. In at least some embodiments, such variability may cause significant degradation in the quality of beamforming.

[0025] Manufacturing-induced variability of the trace length can be eliminated or at least minimized with factory based calibration. For example, the phase array antenna can be placed into an anechoic chamber with auxiliary antennas located in a far field to measure relative phase differences between individual antenna elements. Knowing the relative phase differences, the phase shifts θί can be adjusted to eliminate those relative differences. However, instrumenting the measurement setup for the above procedure and executing the measurements for each antenna can be costly in serial production. [0026] In some embodiments of the present technology, an instrument-less calibration procedure capable of running within short time (e.g., sub-millisecond time) at a boot-up time can characterize manufacturing-induced phase differences between the traces. An additional advantage of the present technology is that, in at least some embodiments, the calibration is run only once for the life of antenna, and the results can be stored on non-volatile memory in the system, e.g., Electrically Erasable Programmable Read-Only Memory (EEPROM), flash memory, disk drive, or other type of non-volatile memory. In some embodiments of the present technology, the calibration method relies on reusing a continuous waveform (CW) generator that is already incorporated into RFIC for the purpose of calibrating receiver (RX) gain and/or IQ (in- phase and quadrature) imbalance. With the inventive technology, a relative calibration can be sufficient, that is, the phase shifts θί can be adjusted/compensated to make Nij's (differences in the length of the traces i and j that were introduced by manufacturing error/tolerances) constant or close to constant for all i,j pairs. Accordingly, the calibration does not have to be absolute, that is, the calibration does not have to result in zero phase offset for each trace. Details of the compensation technology are described below with reference to Figure 5.

[0027] Figures 5A and 5B are schematic illustrations of a transceiver in accordance with an embodiment of the present technology. In an embodiment of the present technology, a phase-locked loop (PLL) of the RFIC can generate a calibration signal CW at, e.g., 8GHz and inject it to a transmitter (TX) antenna. The injected signal can leak over the phase array antenna back to antenna j. In some embodiments, a leakage of about 18-20 dB is acceptable for the calibration. Next, the leakage signal can bypass a mixer circuit in the transceiver 500, and the calibration signal is routed through a coaxial cable back to baseband chip. In some embodiments, the baseband chip includes a complex mixer for frequency within A/D sampling rate. The sampled calibration signal CW can be correlated within the baseband chip to determine the phase shifts for the particular traces. In some embodiments, prior synchronization of baseband digital mixer to RF and baseband PLLs can be used to improve the accuracy of the determination of the phase shifts. The above calibration procedure can be repeated for all traces Ti. Furthermore, the calibration procedure can be extended to tiled phase array antennas by, for example, comparing all phase array antennas to the antenna at the edge of the tile.

[0028] Figure 6A is a graph of signal gain in dB as a function of the AoA for a phase array antenna. Dashed line represents signal gain when traces have equal length, and solid line represents signal gain when the differences in trace length satisfy equation 2. In both cases, the phase array antenna has 32 elements and 5 bit phase resolution, and the trace loss is 0.25 dB/cm. As explained with relation to Figure 2, the traces of equal length assure uniformity of phase shifts of the signal for all traces. However, when all traces have equal length at least some traces, e.g., centrally located traces, are longer than necessary, which causes additional power loss in the centrally located traces. Consequently, the signal gain for the phase array antenna with traces of equal length is smaller than the signal gain for the phase array antenna with traces of unequal length that satisfy equation 2. On average, signal gain of the phase array antenna having unequal trace length is higher for about 1 dB than the signal gain of the phase array antenna having equal trace length. In both cases, the sharp, concentrated peaks represent quantization errors.

[0029] Figure 6B is another graph of signal gain in dB for a phase array antenna having traces of equal and unequal length. Here, the trace loss is 0.5 dB/cm, which is twice as much as the trace loss for the test cases illustrated in Figure 6A. Therefore, the losses associated with the centrally located traces are even higher than in the test case illustrated in Figure 6A. As a result, the signal gain for the equal trace case is about 2.5 dB smaller on average than the corresponding signal gain for the traces that satisfy equation 2.

[0030] From the foregoing, it will be appreciated that specific embodiments of the invention have been described herein for purposes of illustration, but that various modifications may be made without deviating from the scope of the various embodiments of the invention. For example, combinations of phase resolution and fraction of wavelength can be used that are different than those illustrated above. Furthermore, while various advantages and features associated with certain embodiments of the disclosure have been described above in the context of those embodiments, other embodiments may also exhibit such advantages and/or features, and not all embodiments need necessarily exhibit such advantages and/or features to fall within the scope of the disclosure. Accordingly, the disclosure is not limited, except as by the appended claims.

Claims

CLAIMS I/We claim:

1 . A communication system for wireless signals, comprising:

a phase array antenna having a plurality of individual antennas; and

a plurality of electrically conductive traces, wherein individual traces electrically connect corresponding individual antennas with a transmitter, wherein lengths of individual traces Ti, Tk satisfy equation:

Abs ((Ti-Tk) mod (λ)) < λ/Β

where λ is a wavelength of a wireless signal to be applied to the antenna and λ/Β is a fraction of λ.

2. The system of claim 1 wherein individual antennas of the phase array antenna are spaced apart by λ/2.

3. The system of claim 1 , further comprising the transmitter capable of transmitting signal through the traces to the phase array antenna.

4. The system of claim 1 wherein λ/Β is close to zero.

5. The system of claim 1 wherein B is selected from a group consisting of 2, 4, 8 and 16.

6. The system of claim 1 , further comprising a beamformer capable of performing amplitude and/or phase adjustment to a wireless signal to be applied to individual conductive traces.

7. The system of claim 6 wherein the beamformer is an optimal beamformer capable of cancelling phase offsets generated by the unequal trace lengths.

8. The system of claim 6 wherein the beamformer applies complex weights wi that satisfy:

wi = exp (Γ2ττ^"Π/λ) ^* w'i

where w'i is a complex weight that corresponds to the electrically conductive traces having uniform length, and j is unit imaginary number.

9. The system of claim 6, further comprising a Radio Frequency Integrated Circuit (RFIC) capable of sending signals to the beamformer.

10. A communication system for wireless signals, comprising:

a phase array antenna having a plurality of individual antennas;

a plurality of electrically conductive traces, wherein individual traces electrically connect corresponding individual antennas with a transmitter, wherein lengths of individual traces Ti, Tk have unequal trace lengths that cause phase offset errors among the individual antennas of the phase array antenna; and

a beamformer configured to perform amplitude and/or phase adjustment to a wireless signal to be applied to individual conductive traces, wherein the beamformer is an optimal beamformer configured to cancel the phase offsets among individual antennas of the phase array antenna.

1 1 . The system of claim 10 wherein the optimal beamformer applies complex weights wi that satisfy:

wi = exp (Γ2ττΤί/λ) ^* w'i

where w'i is a complex weight that corresponds to the electrically conductive traces having uniform length, and j is a unit imaginary number.

12. The system of claim 10 wherein the lengths of individual traces Ti, Tk satisfy equation:

Abs ((Ti-Tk) mod (λ)) < λ/Β

13. A method for producing a phase array antenna, the method comprising: manufacturing traces for connections to individual antennas of the phase array antenna, wherein the lengths of the traces satisfy Abs ((Ti-Tk) mod (λ)) < λ/Β, where λ is a wavelength of the wireless signal and λ/Β is a fraction of λ; and c

connecting the traces to the phase array antenna.

14. The method of claim 13, further comprising connecting the traces to a transmitter.

15. The method of claim 13, further comprising connecting the traces to the phase array antenna to an optimal beamformer, wherein the optimal beamformer applies complex weights wi that satisfy:

wi = exp (Γ2ττΤί/λ) ^* w'i

16. The method of claim 15, further comprising connecting a Radio Frequency Integrated Circuit (RFIC) to the beamformer.

17. A computer-readable storage medium storing processor-executable instructions that, when executed, cause a beamformer to apply complex weights wi to signals, comprising:

instruction to calculate the complex weights as:

wi = exp (Γ2ττΤί/λ) ^* w'i

where w'i is a complex weight that corresponds to the electrically conductive traces having uniform length, and j is unit imaginary number; wherein lengths of individual traces Ti, Tk that connect the beamformer with elements of a phase array system satisfy equation:

Abs ((Ti-Tk) mod (λ)) < λ/Β

where λ is a wavelength of the wireless signal and λ/Β is a fraction of λ.

18. A method for beamforming, comprising:

emitting a wireless signal from a phase array antenna having a plurality of individual antennas, wherein individual traces that electrically connect individual antennas of the phase array antenna with a transmitter have unequal trace lengths that cause phase offset errors among the individual antennas of the phase array antenna; and

beamforming the RF signal of the phase array antenna, wherein the adjusts amplitude and/or phase of the wireless signal that is applied to individual conductive traces, and wherein the beamformer is an optimal beamformer configured to cancel the phase offsets among individual antennas of the phase array antenna.

19. The method of claim 18 wherein beamforming the RF signal includes applying complex weights wi that satisfy:

wi = exp (Γ2ττ^"Π/λ) ^* w'i

20. The method of claim 18 wherein the lengths of individual traces Ti, Tk satisfy equation:

Abs ((Ti-Tk) mod (λ)) < λ/Β