Calibration Issues of PHY Layer Abstractions for
Wireless Broadband Systems
Antonio Maria Cipriano
∗
, Raphaël Visoz, and Thomas Sälzer
France Telecom Recherche et Developpement, 38-40, rue du Général-Leclerc, Issy Moulineaux, France
e-mail: antonio.cipriano@fr.thalesgroup.com,{raphael.visoz,thomas.salzer}@orange-ftgroup.com
∗
A. M. Cipriano is currently with Thales Communications, Colombes, France
Abstract—Today link-to-system (L2S) interfaces are more and
more used in order to speed up complete system-level simu-
lations. In this paper a simple extension to generic incremental
redundancy (IR) hybrid automatic repeat request (HARQ) strate-
gies is presented for two well-known L2S interfaces: the expo-
nential effective SNR metric (EESM) and the mutual information
effective SNR metric (MIESM). Then we focus on the problem of
the L2S interface tuning, which is necessary to achieve the highest
accuracy of the prediction models. The standard calibration
procedure is compared with a new method, based on the average
(over channel and noise) physical (PHY) layer performance.
The latter, called average calibration procedure, is less time
consuming than the standard procedure. Moreover, we show
by simulation that the optimal calibration factors, calculated
with the two methods, converge to close values thus obtaining
equivalent prediction accuracy for the same L2S interface.
Index Terms—PHY, MAC, PHY abstraction, calibration,
broadband wireless system, OFDM.
I. INTRODUCTION
Multi-cell system-level simulators for broadband wireless
systems need accurate predictions of link-level performance
indicators, like the packet error rate (PER). These predictions
should have low computational cost, in order to let multi-
user/multi-cell scenarios be computationally feasible. At the
same time, they should accurately model the PHY layer be-
haviour, so that new and more complex features of the medium
access control (MAC) layer can be modeled in a precise way
(e.g. fast and adaptive scheduling in time, frequency and space,
adaptive modulation and coding (AMC), HARQ, etc.). The
functional block providing to the MAC those indispensable
PHY layer predictions is called PHY layer abstraction or L2S
interface. A generic L2S model can be rapidly summarized as
follows [1]:
• It collects channel characteristics (fast fading, shadow-
ing, pathloss, interference, etc.) and information on the
resource allocation (power allocation, frequency bands
allocation, etc.).
• It maps the previous data to a set of scalar parameters
(usually one or two) via a function which may depend
on modulation, code type, block size, etc..
• It maps the previously found scalars to a PER prediction.
Previously available L2S interfaces, in use for narrow-band
or code division multiple access (CDMA) systems, are based
on PER curves in additive white Gaussian noise (AWGN)
channel, possibly with correction factors. They make sense
when sent packets experience a unique signal to noise ratio
(SNR). Unfortunately, these predictors are in general far less
effective for new broadband systems based on orthogonal
frequency division multiplexing (OFDM), when very different
SNRs may affect the same data. The same kind of problem is
encountered when using HARQ, e.g. see the GSM/EGPRS
standard [2], which uses a particular IR HARQ protocol.
In order to cope with this problem, new L2S interfaces are
proposed in the literature and a good summary is presented in
[3] (see also [4]).
In this work we focus on two of the most widely used
L2S interfaces, the EESM and the MIESM L2S interfaces
[1]. We propose a simple extension of these predictors in
order to take into account a generic IR HARQ protocol. Other
recent propositions in the literature for HARQ can be found
in [5], [6]. An important issue of L2S interfaces is the so-
called calibration: one or more parameters should be tuned
for providing the highest degree of prediction accuracy. The
interface calibration is a costly procedure based on the outputs
of time-consuming PHY layer simulations. In this paper we
study also the calibration behaviour of the proposed L2S
interfaces and we propose an alternative faster calibration
procedure.
In Sect. II we introduce the standard EESM and the MIESM
L2S interfaces and describe their extension to generic HARQ
protocol. In Sect. III the standard and new calibration pro-
cedure is introduced. In Sect. IV simulations results about
calibration methods are presented. Conclusions are drawn in
Sect. V.
II. L
INK-TO-SYSTEM INTERFACE
A. L2S Interface without HARQ
The L2S interface we deal with compresses the PHY layer
input data into a scalar called effective SNR SNR
eff
(or
effective signal to interference plus noise ratio (SINR) if
interference is take into account) [7]. The effective SNR
summarizes the effect of possible multiple SNRs affecting
the packet by modelling the link as an equivalent AWGN
channel. The map used to derive the estimated PER starting
from SNR
eff
is called in the following AWGN-PER look-up
table (LUT), and it is the PER-performance table obtained for
a given modulation and coding scheme (MCS) under AWGN.
The L2S interface works as follows [8]
978-1-4244-1722-3/08/$25.00 ©2008 IEEE. 1