發表文章

目前顯示的是 2015的文章

MSDU fragmentation / Aggregation MSDU

圖片
With fragmentation the MSDU would be broken into smaller sections and each section encapsulated in an MPDU . Each MPDU is sent in a separate PPDU with the preamble of each PPDU providing a new  channel estimate. (PHY Layer) MSDU fragmented condition:  when its length exceeds a threshold . Each fragment contains an even number of bytes and all fragments are the same size , except the last fragment, which may be smaller. The timing relation between A-MSDU/A-MPDU/PPDU :

SIFS/DIFS/PIFS (802.11 Mac Layer)

圖片
1.What is Short Inter-frame Space (SIFS)? Ans : the small time interval between the data frame and its acknowledgment . Used for the highest priority transmissions enabling stations with this type of information to access the radio link first. 2. What is the content of SIFS ? Ans :This value is fixed per PHY and is calculated in such a way that transmitting station will be able to switch back to receive mode and be capable of decoding the incoming packets. A SIFS duration is a constant value and it depends on the amendments . 3.What is  DCF inter-frame space( DIFS) ? Ans : The DIFS is used by stations operating under the DCF method to transmit data frames and management.(Read more : DCF Method ) 4.How to use DIFS? Ans : A station using the DCF is allowed to transmit if it determines that the medium is idle for the duration of the DIFS, or if it determines that the medium is idle for the duration of the DIFS + the remaining back-off time following the recept...

Random Back-off Procedure (802.11)

圖片
The random back-off count is selected as a pseudo-random integer drawn from a uniform distribution over the interval [0,CW] , where CW, an integer value, is the contention window. Update procedure of CW value 1. CW parameter takes the initial value CWmin . 2. Every unsuccessful MPDU transmit , for example each time an ACK response is not received for a data frame. CW parameter doubles . 3. Every successful MPDU transmit, CW parameter reset . 4. If CW parameter increase due to unsuccessful MPDU transmit and reach the value CWmax , CW parameter keep CWmax until reset . Random back-off procedure 1. As shown in the figure above, STA3 which select the smallest back-off count win the contention and transmit first. 2. STA2 and STA4  suspend their back-off and resume DIFS after the medium goes to idle . 3. STA4, which select the second smallest back-off count win the 2nd contention. 4. And it goes STA2 and finally STA1 transmit. 

Cascade Noise Figure Calculation

圖片
When amplifiers are cascaded together in order to amplify very weak signals, it is generally the first amplifier in the chain which will have the greatest influence upon the SNR because the noise floor is lowest at that point in the chain. If now we have two amplifier with the same gain but different noise figure, which combination of cascading will have more benefit ?  Assume 10dB gain in each amplifier. One amp with NF = 3dB and the other 6dB. When the 3dB NF amplifier is first in cascade, the total noise figure is 3.62 dB. When the 6dB amplifier is first, the total NF is 6.3dB. The amplifier with less noise figure should be cascaded in the first place. If two amplifiers have the same Noise Figure but different gains, the higher gain amplifier should precede the lower gain amplifier to achieve the best overall Noise figure.

IMRR definition on DCR-RF with IQ mismatch

圖片
We define the IMRR(Image-Rejection Ratio) as the ratio of wanted factor to unwanted factor due to IQ-mismatch. The wanted factor is K1 , unwanted factor is K2, thus

Delay Spread and Power Delay profile

圖片
Fourier Transform of Power Delay Profile is called spaced-frequency correlation function, which describes the spreading of a signal in frequency domain . This also give rise to the importance channel parameter - Coherence Bandwidth.

Difference between dB/dBm/dBV

圖片
What is dB ? When a physical quantity , such as power or intensity is measured relative to reference level , it can be expressed in decibels(dB). Decibels is a dimensionless unit . For example, the Signal-To-Noise Ratio can be represent in dB. What is dBm ? dBm stands for an absolute power level , which we can define as 10*log(P/1mW) , where the value of P is power in watts . So P=1mW equals 0 dBm. What is the relation between dBm and dBV ? dBm = dBV - 10log(R) + 30 in a system with resistance R. Assume R = 50 ohm, then the relation between dBm and dBV can be describe as dBV = dBm -13. What is dB digital code word ? Assume we have N bits digital code word, then dB digital code word can be define as the following equation :

Noise Equivalent Bandwidth

圖片
An amplifier or a filter will have a non-ideal frequency response , in order to easily describe the noise power effect through the amplifier or a filter, we need a Noise Equivalent Bandwidth.

S-Parameter and VSWR

圖片
S-parameters describe the input-output power relationship between 2 ports in an electrical system. Notation : for example , Port 1 and Port 2, S 12  represents the power transferred from Port 2 to Port 1. What is the physical meaning of S 21   = - 10dB ? It means if we have 0dB power is delivered to Port 1, then we only have -10dB of power is received at Port 2. What is the physical meaning of return loss/reflection coefficient ? We denote the return loss as S 11 , which represent how much power is reflected from the Port 1/Antenna 1. For example S 11 = 0dB means that all the power is reflected, nothing is radiated. Base on the former description, we can easily observe from a frequency response plot of |S 11 | to decide which frequency will be efficient, say in 2.5 GHz in the figure above. What is the meaning of VSWR ? Impedance mismatch will induce return loss, and this cause a "reflected voltage wave" , which creates standing wave along the transmission...

Delta-Sigma ADC

圖片
For most kinds of ADCs : data rate equals sampling rate ,each input sample converts to one output code. Delta-sigma ADC : acquires many input samples to produce one output code.  Which implies  Delta-sigma sampling algorithm lengthens the acquisition time. In the above figure ,the delta-sigma’s modulator samples and quantize the input signal in a coarse fashion, by producing a one-bit stream of data at a very high rate. Delta-sigma modulator includes an integrator , which has the effect of shaping the quantization noise. The F d value is the corner frequency of the internal digital/decimation filter. One uses F d to define the corner frequency of delta-sigma converter’s external anti-aliasing filter . Observe the figure A and figure B above. The modulator sample rate (Fs) shapes the quantization bandwidth. The data rate (F d ) is always smaller than Fs, as in  Figure A and B . Signals in the i...

Rayleigh Fading and Rician Fading

圖片
Rayleigh fading is a reasonable channel model when there are many objects in the environment that scatter the radio signal before it arrives at the receiver. The central limit theorem holds that, if there is sufficiently much scatter , the channel impulse response will be modeled  as a Gaussian process irrespective of the distribution of the individual components. The Rician K factor, defined as the ratio of signal power in dominant component over the scattered, reflected power , which shows in the below figure  is a easy way to understand the behavior of a short-range wireless channel. Short conclusion :  1.  Rayleigh fading is most applicable when there is no dominant propagation along a line of sight between the transmitter and receiver .  2. Rician fading is used while there is a dominant line of sight . (Which is usually used in short-range wireless channel)

Moving Average and Exponential Average

圖片
A block moving Average have the following equation : The actual implementation will require : (1) N-tap delay element (2) 1-tap delay element (3) 1 shifter (If N = 2 n ) (4) 2 adder An exponential Average will have the following equation : Alpha value will decide the convergence speed and the weight of the current input signal or data. If the alpha value are smaller then the weight of the current input signal affect less. The actual implementation will require : (1) 1-tap delay element. (2) 2 shifter (If alpha = 2 n ) (3) 1 adder We can easily find out that the Exponential Average are easily to implement than the Moving Average. But if we need to observe the instantaneous change data , the moving average method will be accurate than exponential average method.    

Spread Spectrum Clock Generation (SSCG)

圖片
Spread-spectrum clock generation (SSCG) is used in some  synchronous digital system   to reduce the spectral density of the  electromagnetic interference (EMI).   Jitter will cause transmitted bit error Because the generated clock will have variation between 5 GHz ~ 4.97 GHz, thus in the time domain , the clock waveform will have time variation. If the variation is too much, then the received waveform will have certain error because of clock timing variation which can be founded in the following figure :

Perl Learning Note - Column Swap Function

Suppose we want to have a function to deal with a global matrix (@matrix) to do column swap, then we can declare the  function  as following : sub ColSwap {     my ($Idx1,$Idx2) = @_;  # Idx1 and Idx2 are global variable     my ($idx, $Tmp0);       # idx and Tmp0 are local variable , only exist in this function     for ($idx = 0; $idx < $max_col; $idx++)     {            $Tmp0 = $matrix[$Idx1*$max_col + $idx];            $matrix[$Idx1*$max_col + $idx] = $matrix[$Idx2*$max_col + $idx];            $matrix[$Idx2*$max_col + $idx] = $Tmp0;     } } NOTE : matrix and max_col are global variable , we don't need to give any input argument to ColSwap. We can use this kind of function using  &ColSwap (Column#1 , Column#2); We can return value like we  program  in C , for example : s...