A muon detector has to fulfil four tasks:

• Muon identification: this is done through absorption of other long-lived particles in the calorimeters (made of high-Z material) and through decay of short-lived particles on their path, before reaching the outside muon system.

• Muon charge determination: by means of the bending direction in the magnetic field (positively and negatively charged muon bend in opposite direction). The magnetic field in the muon barrel region is below 1 T and rather uniform. The charge determination has to be better than 95%.

• Muon momentum determination: is done through determination of the bending radius. The stand-alone muon system can provide 10% momentum resolution which can be improved by combining the muon track segments with those from the high-resolution tracker. The momentum resolution is limited by multiple scattering in the return-yoke iron and the materials up-front up to about 1 TeV energies. At higher muon energies, the chamber resolution of 100 micron starts to limit.
• Trigger information: In hadron environments such as the LHC, the selection of the interesting events over a 10 orders of magnitude higher background is the main challenge. This selection is done by the trigger which looks for significant kinematic decay patterns or final state particles. Since muons are often present in the final state, a fast and reliable muon identifaction will constitute the first level muon trigger.

The CMS barrel muon chambers are big drift-chambers (up to 4 x 2 qm cross-section and several hundred channels per chamber). A through-going charged particle creates electrons and ions which drift in the elctric field. Electrons drift to the anode wire in the centre of the cell, while the slower ions drift to the cathode. The ion signal is not read out. The electric field is uniform over large ares to allow a constant drift velocity, which is 55 microns per nanosecond in our case. Near the wire the field increases by several orders of magnitude and electron multiplication occours. About 10^5 electrons arrive at the wire and will be read out. A muon detector consists of several layers, two times four in the important r-phi-plane and one times four in the orthogonal direction. Individual cells are rectangular with dimensions of 4 x 1 cm.The layers are stacked up from aluminium plates spaced by so-called I-beams (the picture shows where the name comes from). These I-beams are covered by a conductive foil constituting the cathode. In the cell-centre is a 50 micron wire strung with 3 N tension, constituting the anode.
The rectangular shape of the cell requires field shaping in the middle region to avoid that field lines leave the cell. Electrons following these field lines would be lost and the signal reduced. Strips of conductive foil are glued to the aluminium plates whose electric field form the elctric field as seen in the picture. The chambers are operated with voltages of 3.7 kV for the wires, -1.2 kV for the cathodes and 1.8 kV for the field shaping electrodes. The amplification gas is ArCO_2.