Fnu ISHAQ BASHA ZAKIR AHMED

25 years of experience in design, research, and deployments of wireless communication, multimedia signal processing algorithms including RF system integration into consumer electronic products and RF instrumentation

https://scholar.google.com/citations?user=A5mQHUUAAAAJ&hl=en

[1] I. Zakir Ahmed, H. R. Sadjadpour and S. Yousefi, "An Optimal Low-Complexity Energy-Efficient ADC Bit Allocation for Massive MIMO," in IEEE Transactions on Green Communications and Networking, vol. 5, no. 1, pp. 61-71, March 2021, doi: 10.1109/TGCN.2020.3039282.

[2] I. Zakir Ahmed, H. R. Sadjadpour and S. Yousefi, "An Information-Theoretic Branch-and-Prune Algorithm for Discrete Phase Optimization of RIS in Massive MIMO," in IEEE Transactions on Vehicular Technology, vol. 72, no. 6, pp. 7395-7410, June 2023, doi: 10.1109/TVT.2023.3237682.

[3] I. Zakir Ahmed, “Down Scaling of Images to Arbitrary Sizes Using Fractional Filtering in Spatial Domain on Independent Image Blocks of Arbitrary Size”, IEEE International Conference on Signal and Image Processing, Hubli, India. [ISBN -10:0230-63005-7, ISBN-13: 978-0230-63005-5]

[4] I. Zakir Ahmed, Vijaya Yajnanarayana “A Time Domain Based Efficient Block Decision Algorithm for Audio Coders” 7th IEEE International Symposium on Communications and Information Technologies. Sydney Australia. [ISBN: 1-4244-0977-2]

[5] I. Z. Ahmed, H. R. Sadjadpour and S. Yousefi, "Information-Assisted Dynamic Programming for a Class of Constrained Combinatorial Problems," in IEEE Access, vol. 10, pp. 87816-87831, 2022, doi: 10.1109/ACCESS.2022.3198964.

[6] I. Zakir Ahmed, Krishna Bharadwaj, Ramesh Krishnan, Vijaya Yajnanarayana, "Estimation of Sample Clock Frequency Offset Using Error Vector Magnitude", US Patent Number: 8442161. Grant Date: 14th May 2013

[7] I. Zakir Ahmed, Vijaya Yajnanarayana, Krishna Bharadwaj, “Blind mechanism for the joint estimation of frequency offset and phase offset for QAM modulated signals”, US Patent Number: 8446974. Grant Date: 21st May 2013

[Apple Inc.]

- Implemented of crystal calibration for UWB front end. The solution involved identifying two capacitor values (C1,C2) which can take 128 discrete value each. The Newton's method of gradient decent was modified to take into account the non-convexity of the error profile as function of (C1,C2). This resulted in a very tight clock within +/-0.1ppm of error thereby significantly improving UWB ranging performance.

- Implemented a Sobel-filter based image processing technique to improve the SNR of the range-doppler plots of a typical FMCW radar.

- Implemented a reinforcement-learning based algorithm to identify targets "behind-the-wall" for a UWB pulsed-based radar at sub6Ghz frequency ranges.

- Characterizing UWB wireless channels during deep-fade and normal-mode of operations using deep-learning frameworks.

Experimental campaign to evaluate the Angle-of-arrival of the target under consideration for a UWB pulsed radar.

[National Instruments Inc.]

[WLAN 802.11ac/ax] :

- Implementation of RF IQ impairments estimation and correction algorithms 

This included blocks like Carrier frequency offset correction, phase tracking algorithm, IQDC offset estimation, IQ Imbalance estimation for WLAN 802.11ac/ax.

- Implementation of Beam-forming Calibration for NI-VST using Chirped Pulse with LabVIEW:

Implemented a mechanism to estimate the delay (and hence the phase difference) between two chirped pulses generated from 2 VSTs tightly synchronized in time and using a shared Local Oscillator (LO) for a given experimental setup and cable configuration. The intention was to estimate the time offset and hence measure calibration delay (Phase) for Implicit beam-forming measurement for 801.11n/ac signals. These calibration correction matrices were finally written down into ROM of Broadcom BCM 4360/4366 to measure the BF gain.

- Algorithm implementations for improvising end-user experience:

- Implementation of a Cyclic Prefix based Joint Carrier Frequency Offset (CFO) and Sample Clock Offset (SCO) Estimation for 802.11ac

- Implementation of 1024 QAM Modulator and Hard decision demodulator to support coding scheme (MCS-10 and MCS-11) within 802.11ac standard.

-  An optimal Polyphase Filter Implementation for Sample Rate Conversion for rational factors.

- Implementation of Sample Rate Conversion for rate change factors being irrational.

- Implementation of Noise Figure Measurement for High Gain Receiver chains

[5G-NR] : Integration of NI’s measurement framework (RFmx) wrapped in Matlab into customer bench for performing LTE and 5G-NR DPD and ET measurements.

-  Design and prototype of Client-Server architecture with python based sequencer running on a client on macOS and NI Instrument running Matlab wrapped RFMx on windows as Server.

- Integration of NI’s SCPI framework for mmWave 5G VST into customer test bench to test DPD for 28G mmWave PA.

[Bluetooth] : Designed and implemented the Bluetooth Enhanced Data Rate (EDR) In-band Emission spurious Emission (IBE) Measurement. This measurement was developed as per the Bluetooth test specification v1.2/2.0/2.0+EDR/2.1/2.1+EDR/3.0/3.0+HS/4.0. The design mimics a swept spectrum analyzer with a configurable Resolution Bandwidth (RBW) filter, Video Bandwidth (VBW) filter, Hold type, Detector Type and number of trace points on a Vector Signal Analyzer (VSA). This measurement was tested with different Bluetooth packet types like 2-DH5, 2-EV5 etc. and measurement results were compared with the swept signal analyzer.

[WCDMA/HSPA+] :

Added waveform Generation and Analysis Features for NI WCDMA/HSPA+ Toolkit. Features added include:

-  Auto Detection of Modulation Type and Channels present in the WCDMA/HSPA+ Signal.

-  Large range CFO estimation and correction for WCDMA/HSPA+ signals

-  Added Higher order Demodulation for HSPA+ channels [16QAM, 64QAM on DL and  4-PAM on UL]

-  Added HPSA+ MIMO spatial multiplexing capabilities into toolkit

[GSM/EDGE/EDGE+] :

- Designed a cascaded LMS Equalizer followed by a Decision Feedback Equalizer based detector for  Performing GSM Phase Error and EDGE/EDGE+ EVM Measurements.

- Implemented EDGE-MCS-4 and MCS-5 Channel coding for Downlink PDTCH channel as part of Non Signaling Tests (Reduced Signaling) for Qualcomm DUT. Tested the same using QRCT (QDART tools)

[LTE-Advanced, LTE] :

-  Review of Adjacent Channel Power Measurement for LTE-Advanced Release 10 for Aggregated Carriers.  Spectral Emission Mask Measurement for the same. Modulation Accuracy and Evaluation of Parameters like composite Error Vector Magnitude (EVM), Subcarrier EVM, In Band Emission (IBE) and Spectral Flatness (SF) for LTE-Advanced.

-  A Novel EVM Evaluation technique was designed and implemented which resulted in significant improvement in test times. This method was patented [1], mentioned above.

[CDMA2K and TD-SCDMA] :

-  Implementation of CFO Estimation algorithm for Uplink RC1 CDMA2K for EVM Measurements

- Implementation of Midamble code and Shift detection in TD-SCDMA uplink Signal.

- Optimization of TD-SCDMA uplink Demodulation for EVM Measurements.

[Signal Intelligence] : 

- Design and implementation of a customized Blind CPM (Continuous Phase Modulation) transreceiver that can correct Doppler’s upto +/-20Khz for 1MHz bandwith Signals.

- Automatic Identification of the Modulation Type, Symbol Rate (In case of a digital scheme) of a  completely unknown signal. (Added 2FSK, 4FSK and OQPSK classifiers in the existing legacy code)

[PA Characterization and Linearization] :

-  AMAM/AMPM Measurement: Designed and implemented the AMAM/AMPM Measurement using LabVIEW on the NI-PXI-5663 6.6Ghz Spectrum Analyzer. The Measurement made use of a special technique to compute the AMAM and AMPM Characteristics of the Power Amplifier by using the stored reference signal. This Measurement can take any standard signal like WCDMA/LTE/WLAN etc. as the input and return the P1dB, Average Gain of the PA, in addition to curve fit polynomial coefficients of the AMAM and AMPM Characteristics.

- Digital Pre Distortion (DPD): Designed and implemented a memoryless Pre Distortion Technique for a WCDMA/LTE signals on a NI-PXI 5673, 6.6Ghz Signal Generator. The method employs a simple characteristics inversion technique to achieve Pre Distortion.

- Envelope Tracking (ET): Designed and implemented a delay computation Measurement on the NI-PX 5663 6.6Ghz Vector Signal Analyzer and NI-PXI 5644R Vector Signal Transceiver. Developed and implemented a technique that could estimate the delay in range of 50nsec and above, with an accuracy of +/-5nsec.

[Motorola Inc.]

[Image Processing] : Image Resizing for iHP-120 Mulitcodec Jukebox: One man team involved in defining the problem for image downsizing to fit display of the iHP-120 jukebox from iRiver. The problem was extended to downsize the images to fit display of arbitrary sizes (With Memory and MHz constraints). This effort, lead to design of a novel algorithm, its design and implementation on Coldfire platform, integration into the iHP-120 jukebox and also involvement in the support phase. This algorithm was published at IEEE conference on Signal and Image Processing.

[Audio signal Processing] : 

- Design and implementation of the psychoacoustic models for MP3 and AAC encoders. The blocks include time-to-place (human ear’s frequency-response) mapping from FFTs, tone-masking-noise, noise-masking-tone experiments, tonality index, and processing-block length identification [4].

- Implementation of 32-subband filter bank for MP3 as well as Modified Discrete Cosine Transform (MDCT) in AAC and its mathematical optimizations to reduce computational complexity.

- Matlab prototyping to floating-point C conversion of the entire codec end-to-end. Fixed point design having the target DSP (or microcontroller) in mind. Assembly coding of key mathematically rigorous modules. The target DSPs/microcontrollers where the MP3 and the AAC codes were deployed are Motorola Coldfire (MCF5240 family),  ARM9/ARM11, and Motorola Starcore (SC140E core).

- As technical leader involved in identifying the performance parameters, requirements gathering, customer interface, design and development of algorithms, Fixed point design, and assembly coding on ARM9/ARM11 for MP3 decoder, WMA-9 decoder and AAC-LC encoder.

- One man team involved in Algorithmic analysis of Audio Image Expansion module for a 5 band Graphic equalizer for mobile phone application. Also ported this algorithm onto the tensilica platform.

[Speech signal Processing] : Fixed point algorithm design and implementation of Fixed point design and development of algorithm for G728 Low-delay speech codec.  Assembly coding of the codec onto SC140 platform. Also published a research paper and a patent discloser on this effort.

[2G/3G Modems] :

- Implementation of Maximum Likelihood Channel Equalizer in Assembly Language of Motorola DSP SC140E (Starcore) for GSM modems. The solutions implemented on SC140E was repurposed over to the TI DSP C55x+.

- Implementation of Symbol rate Chain for W-CDMA/HSPA+ wireless infrastructure systems. Assembly coding of convolution encoder and Viterbi decoding modules on SC140E platform for 3GPP modems.

• Hamid Sadjadpour, Professor and Graduate Director of Electrical Engineering, University of California, E2-245B, University of California, Santa Cruz, CA, 95064, hamid@soe.ucsc.edu
• Vijaya Parampalli Yajnanarayana, Master researcher, Ericsson research, Bangalore, vijayapy@gmail.com
• Craig E. Rupp, Director of Engineering and Senior Fellow, The Climate Corporation, Chicago, IL, ruppinator@gmail.com