Commius' relief force was, says Caesar, less than a mile from the Roman line--Alesia is a hill city in the middle of a rather vast plain, and it's pretty easy to see stuff from the top of the hill way out in the distance. According to Caesar, when Commius pitched his camp outside the Roman lines the Gauls within the city came out, ready for a sally, and filled in one of his trenches in preparation for an assault on the Roman position. There were several coordinated attacks, which didn't all proceed the same way. Once Commius had taken his position he sent out light infantry and cavalry in preparation for a larger sally, while Vercingetorix's Gauls prepared to assault the Roman works as soon as Commius' force was clear to move forward--Caesar's German cavalry swept the Gallic cavalry back, however, and the assault was broken off. The Gauls waited a day to prepare grappling hooks and stuff, and tried a night assault. The Gauls within the city, like troops in every town under siege, were largely under arms the whole time, especially since they were awaiting a sortie from their relief army the whole time--Caesar is quite clear on this last point:

seque ad eruptionem atque omnes casus comparant.

Caesar says that Commiu' men approached the Roman works under cover of darkness, then let out a great shout to signal to the Gauls within:

Subito clamore sublato, qua significatione qui in oppido obsidebantur de suo adventu cognoscere possent

and that Vercingetorix signaled his men with a trumpet-call:

Eodem tempore clamore exaudito dat tuba signum suis Vercingetorix atque ex oppido educit

At the final assault, Commius had reconnoitered the area and found a weak spot in the Roman lines. Vercingetorix observed Commius' approach from the citadel of the city:

Vercingetorix ex arce Alesiae suos conspicatus ex oppido egreditur

In any case, ancient armies had a number of different ways to communicate and signal. The most important were smoke-signals and fire signals, the former being used during daylight and the latter at night. Especially in Greek literature beacon-fires are well-attested--Clytemnestra is awoken in the Agamemnon when a sentry spots the beacon signalling Troy's fall. Herodotus and Xenophon both speak of shields being used for signals--Herodotus has a traitor signal the Persian fleet after Marathon by flashing his shield in the light from Pentelicus, and Xenophon describes Lysander as ordering his scout ships to raise a shield to signal to his main fleet when it was safe to attack the Athenians at Aegospotomi. Torch signals are also well-attested--my favorite instance is Thucydides' description of the Plataeans successful confusion of the Spartan and Theban torch signals raised to alert the troops that the Athenians were withdrawing from under their noses. Polybius (X.44) quotes a lengthy passage of Aeneas Tacticus' where he speaks of a rather complex method of fire signaling, within a passage devoted entirely to discussion of fire signals, which he says are "μεγίστας δὴ παρεχομένου χρείας ἐν τοῖς πολεμικοῖς, ἀνεργάστου πρότερον ὑπάρχοντος."

We should also not suppose that Commius and Vercingetorix had no direct communication. Though Caesar does not appear to think that any messengers got through his lines (and it's likely that he's right), it is possible. Some parts of Caesar's lines were quite thin, otherwise the third assault would never have gotten so close to success. Caesar himself says:

Romanorum manus tantis munitionibus distinetur nec facile pluribus locis occurrit.

so it's entirely possible that some of Commius' men might have made it within the city to meet with Vercingetorix, though I don't think it's likely