Re: [Discuss-gnuradio] Question regarding gr_mpsk_receiver_cc::mm_error_tracking

[Kyle Zhou]

Hi Ian,
I think the code is correct. Eqn. (8) in the referred paper is wrong. 
However, Fig. 1 in the paper is right, and the gnuradio code is exactly 
the same as Fig.1.
Kyle
> Date: Wed, 7 Apr 2010 13:51:10 +0930
> From: "Ian Holland" <address@hidden>
> Subject: [Discuss-gnuradio] Question regarding
>         gr_mpsk_receiver_cc::mm_error_tracking
> To: <address@hidden>
> Message-ID:
>         <address@hidden>
> Content-Type: text/plain;       charset="us-ascii"
>
> Hi All
>
> I am trying to understand how the optimised modified Mueller and Muller
> algorithm is implemented in GNU Radio.
>
> I had a look at the method gr_mpsk_reciever_cc::mm_error_tracking, to
> see how this is done. As far as I can tell, lines 242-245 are intended
> to implement equation (8) of the referenced paper, where mm_error
> corresponds to mu(k) in eqn. (8). However, if I have interpreted this
> correctly, what is implemented is actually:
>
> \mu(k) = Real{[p(k) - p(k-2)] \times \hat{c}^{*}(k-1) - [\hat{c}(k) -
> \hat{c}(k-2)] \times p^{*}(k-1)},
>
> whereas eqn. (8) in the referenced omM&M paper, is actually:
>
> \mu(k) = Real{[\hat{c}(k) - \hat{c}(k-2)] \times p^{*}(k-1) +
> \hat{c}^{*}(k-1) \times [p(k) - p(k-2)]}
>
> Have I missed something here? Are these lines of code not meant to
> implement eqn (8) as I suspected?
>
> Thanks
>
> Ian.
>