You seem to be asking how a paper kanji dictionary is indexed.
They are usually ordered by radical and stroke count.
This is one reason why the ability to count strokes is important
A radical is a particular type of sub component that makes up a kanji. There are 214 radicals, many (most?) of them characters in their own right. A number of them are written in several ways.
for instance radical 86 火 hi also appears as 灬 rekka and hihen (the right hand side of 灯)
so the kanji 照 in 照明[しょうめい] is indexed under 火 and has 4 strokes(the radical) plus 9strokes (the rest). This should be enough to find it in a dictionary.
(Although if you don't realise that 火 is the radical you might mistakenly try to search under 日. recognising radicals is another skill in using a dictionary)
Some dictionaries also have on and kun indexes to allow you to find a kanji you know the reading for (beware the many homonyms )
Although it isn't ordered by radical and has its own quirky system The Kodansha Kanji Learners Dictionary is worth a look.
Also for how kanji work Lets Learn Kanji by Mitamura & Mitamura is interesting.
Personally I find computer based searches quicker and more useful. Either stand alone solutions like JEDict or online solutions like http://www.jisho.org/
And there are also solutions that allow kanji handwriting recognition such as a NintendoDS kanji dictionary, or iPhone/iPodTouch Chinese Character recognition (for fingerpainters), or the later models of Wordtanks
(PS In lists of kanji they are usually ordered by Onyomi in あいうえお order.)